Append PII to any phone.

What each endpoint returns

Quickstart

Three steps to your first call.

1. 1Sign up at talkdatatome.online/signup. Any plan works to start.
2. 2Create a key on the Developers page. Copy it right away. The key is shown once and starts with rk_.
3. 3Make a call with the key in the X-API-Key header. Pick your language below.

Prefer to skip straight to a running call? The API Playground runs every endpoint from your browser with your key, shows real latency, and hands you the exact cURL to reproduce it.

curl -X POST https://www.talkdatatome.online/api/v1/resolve \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"identifier": "4155550142"}'

import requests

resp = requests.post(
    "https://www.talkdatatome.online/api/v1/resolve",
    headers={"X-API-Key": "rk_your_key_here"},
    json={"identifier": "4155550142"},
)
print(resp.json())  # {"matched": true, "first_name": "...", ...}

const resp = await fetch("https://www.talkdatatome.online/api/v1/resolve", {
  method: "POST",
  headers: {
    "X-API-Key": "rk_your_key_here",
    "Content-Type": "application/json",
  },
  body: JSON.stringify({ identifier: "4155550142" }),
});
console.log(await resp.json()); // { matched: true, first_name: "...", ... }

That is the real-time /resolve endpoint, which answers a single identifier in well under a second. To append data to a whole list at once, use the batch /enrich endpoint below. Both take the same key.

Authentication

Every request must include your secret key in the X-API-Key header. Keys look like rk_… and are shown to you once at creation. Treat your key like a password and never expose it in client-side code. A missing, invalid, inactive, or expired key returns 401.

Create and manage keys yourself on the Developers page: issue a new key, pause or revoke any key, and watch this month's usage. The same key works for every endpoint on this page.

X-API-Key: rk_your_key_here

OpenAPI spec

The full contract is published as an OpenAPI 3.1 document covering every key-authenticated endpoint on this page, with request and response schemas. Load it into Postman or Insomnia, or point a client generator at it to get a typed SDK in your language. No key is needed to read it.

Open the OpenAPI spec

https://www.talkdatatome.online/api/openapi.json

The spec is generated from the running code, so it always matches the deployed behavior. Signed in, you can also reach it any time from API in the top nav.

Enrich

Send one phone via {"phone": "..."} or up to 100 via {"phones": ["...", "..."]}. Phones can be in any common format. We normalize to 10 digits and de-duplicate before resolving. Every submitted phone comes back in results, matched or not.

curl -X POST https://www.talkdatatome.online/api/v1/enrich \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"phones": ["+1 (415) 555-0142", "2025550173"]}'

{
  "phones": [
    "+1 (415) 555-0142",
    "2025550173"
  ]
}

{
  "requested": 2,
  "matched": 1,
  "results": [
    {
      "phone": "+1 (415) 555-0142",
      "matched": true,
      "confidence": "high",
      "verified": false,
      "matched_by": "phone",
      "phone_people": 1,
      "first_name": "Jordan",
      "last_name": "Avery",
      "age": 47,
      "gender": "M",
      "marital_status": "Married",
      "homeowner": "Home Owner",
      "household_income": 112500,
      "address": "1240 Oak Ridge Dr",
      "city": "San Francisco",
      "state": "CA",
      "zip": "94110",
      "mobile_phone": "4155550142",
      "landline_phone": "4155550199",
      "email": "jordan.avery@example.com",
      "ip_address": "73.92.18.204",
      "source_url": "homeimprovementquotes.com"
    },
    {
      "phone": "2025550173",
      "matched": false
    }
  ]
}

Null fields are omitted from a matched result, so a record only carries the attributes we actually hold. An unmatched phone returns just {"phone": "...", "matched": false}. You are billed for matched rows only.

{
  "phones": ["5303219112"],
  "min_confidence": "high"
}

{
  "requested": 1,
  "matched": 0,
  "results": [
    {
      "phone": "5303219112",
      "matched": false,
      "reason": "below_min_confidence",
      "confidence": "low"
    }
  ]
}

curl -X POST https://www.talkdatatome.online/api/v1/enrich \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"records": [{"phone": "4045550142", "email": "jane.doe@example.com", "first_name": "Jane", "last_name": "Doe", "address": "123 Main St", "city": "Atlanta", "state": "GA", "zip": "30301"}]}'

{
  "records": [
    {
      "phone": "4045550142",
      "email": "jane.doe@example.com",
      "first_name": "Jane",
      "last_name": "Doe",
      "address": "123 Main St",
      "city": "Atlanta",
      "state": "GA",
      "zip": "30301"
    }
  ]
}

{
  "requested": 1,
  "matched": 1,
  "results": [
    {
      "matched": true,
      "matched_by": "phone",
      "confidence": "high",
      "verified": true,
      "first_name": "Jane",
      "last_name": "Doe",
      "address": "123 Main St",
      "city": "Atlanta",
      "state": "GA",
      "zip": "30301",
      "homeowner": "Home Owner",
      "is_homeowner": true,
      "line_type": "mobile",
      "dnc": 0,
      "age": 47,
      "gender": "F",
      "income_band": "H. $75,000-$99,999",
      "income_midpoint": 87500,
      "credit_range": "C. 700-749"
    }
  ]
}

Response fields

That is the default response. Every dataset in the graph, with its fields, types, and definitions, is documented in the data dictionary. No login needed, so it is safe to send to your team.

Field groups: the full data dictionary on batch enrich

curl -X POST https://www.talkdatatome.online/api/v1/enrich \
  -H "X-API-Key: YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"phones": ["4155550142"], "fields": ["demographics", "property", "auto", "executive"]}'

{
  "requested": 1,
  "matched": 1,
  "results": [
    {
      "phone": "4155550142",
      "matched": true,
      "first_name": "Jordan",
      "last_name": "Avery",
      "age": 47,
      "gender": "M",
      "marital_status": "Married",
      "homeowner": "Home Owner",
      "household_income": 112500,
      "address": "1240 Oak Ridge Dr",
      "city": "San Francisco",
      "state": "CA",
      "zip": "94110",
      "mobile_phone": "4155550142",
      "landline_phone": "4155550199",
      "email": "jordan.avery@example.com",
      "ip_address": "73.92.18.204",
      "source_url": "homeimprovementquotes.com",

      "demographics": {
        "age": 47,
        "birth_year": 1979,
        "credit_range": "B. 750-799",
        "credit_midpts": 775,
        "dwelling_type": "Single Family",
        "length_of_residence": 9,
        "children_hh": 2,
        "child_aged_7_9_hh": "Y",
        "occupation_category": "Management",
        "education": "College",
        "net_worth_hh": "$250,000-$499,999",
        "home_improvement": "Y",
        "diy": "Y",
        "cooking": "Y",
        "dog_owner": "Y",
        "camping_hiking": "Y",
        "travel_affinity": "Y",
        "investments": "Y",
        "credit_card": "Y",
        "catalog_affinity": "Y",
        "veteran_hh": "N",
        "donor_affinity": "Y"
        // ... ~108 more attributes on a real match (nulls omitted)
      },

      "property": {
        "yr_built_orig": 1987,
        "living_sqft": 2140,
        "land_sqft": 6300,
        "nmbr_bedrooms": 4,
        "tot_baths_est": 2.5,
        "full_baths": 2,
        "half_baths": 1,
        "pool": "Y",
        "owner_occupied": "Y",
        "estimated_value": 918000,
        "assess_val_prop": 742000,
        "sale_date": "2017-08-14",
        "sale_amt": 685000,
        "county_name": "San Francisco"
        // ... ~150 more property fields on a real match
      },

      "financial": {
        "discretionary_income": 38500,
        "mortgage_refinance_intent": "Y",
        "mortgage_open1st_intent": "N",
        "automotive_loan_intent": "N",
        "bank_card_intent": "Y",
        "personal_loan_intent": "N",
        "card_offer_response": "Y",
        "mortgage_offer_response": "N"
      },

      "vehicles": [
        {
          "vin": "SAMPLEVIN0000TEST",
          "make": "Toyota",
          "model": "Tundra",
          "year": 2019,
          "fuel": "Gasoline",
          "msrp": 41500,
          "body_type": "Truck",
          "class": "Mainstream",
          "drive_type": "4WD",
          "engine_cylinders": 8,
          "transmission_type": "Automatic",
          "purchase_date": "2021-03-02",
          "purchase_new": "N",
          "number_of_vehicles_in_hh": 2
          // ... full vehicle spec on a real match; up to 5 vehicles
        }
      ],

      "executive_roles": [
        {
          "company_name": "Avery Roofing Co",
          "standardized_title": "Owner",
          "executive_level": "C-Level",
          "executive_department": "Executive",
          "primary_sic_description": "Roofing Contractors",
          "location_sales_description": "$1M-$2.5M",
          "location_employee_description": "10-19",
          "bus_address_city": "San Francisco",
          "bus_address_state": "CA",
          "bus_phone": "4155550171"
          // ... ~100 more executive/business fields; up to 3 roles
        }
      ]
    }
  ]
}

Error codes

Errors return a JSON body shaped like {"error": "..."} with the matching HTTP status.

Rate limits

List pulls by ZIP

curl -X POST https://www.talkdatatome.online/api/v1/pull   -H "X-API-Key: YOUR_KEY"   -H "Content-Type: application/json"   -d '{
    "zips": ["77066", "77069", "77070"],
    "max_records": 5000,
    "filters": {
      "min_home_value": 150000,
      "year_built_on_or_before": 2010,
      "phone_requirement": "mobile"
    }
  }'
# → 202 {"queued": true, "order_id": "...", "list_id": "ZIPLIST-...",
#        "cancel_url": "/api/v1/pull/<id>/cancel", "billing": {...}}

# ZIP-3 prefixes cover a whole metro in one entry instead of listing every ZIP:
curl -X POST https://www.talkdatatome.online/api/v1/pull   -H "X-API-Key: YOUR_KEY"   -H "Content-Type: application/json"   -d '{"zips": ["331", "332"], "max_records": 5000, "filters": {"phone_requirement": "none"}}'
# → 202 {"queued": true, "order_id": "...", "list_id": "ZIPLIST-...", "zips": 2, ...}

# Poll status + get the download link. Pages 20 at a time by default:
curl "https://www.talkdatatome.online/api/v1/pull?limit=20&offset=0" -H "X-API-Key: YOUR_KEY"
# → {"orders": [{"id": "...", "status": "completed", "record_count": 5000,
#                "download_url": "/api/v1/pull/<id>/download"}],
#    "limit": 20, "offset": 0, "total": 353, "has_more": true}

# Changed your mind? Cancel before it is delivered:
curl -X POST https://www.talkdatatome.online/api/v1/pull/<id>/cancel -H "X-API-Key: YOUR_KEY"
# → {"canceled": true, "billed": false, "delivered": false}

Bulk appends

const rows = phones; // your full list
for (let i = 0; i < rows.length; i += 100) {
  const batch = rows.slice(i, i + 100);
  const resp = await fetch("https://www.talkdatatome.online/api/v1/enrich", {
    method: "POST",
    headers: { "X-API-Key": process.env.TDTM_KEY, "Content-Type": "application/json" },
    body: JSON.stringify({ phones: batch }),
  });
  const { results } = await resp.json();
  // append results to your output as you go
}

Common recipes

The Resolution API is the fast, single-identifier twin of the Enrichment API. Send any one identifier and get a small response back fast enough to sit inside an inbound-call routing loop. It is built for real-time bidding (RTB) pre-qualification, on-the-fly enrichment, and pixel resolution. Where /api/v1/enrich batches up to 100 phones against the warehouse, this endpoint answers a single lookup from a hot in-memory store, so it answers fast enough to sit inside a live call flow (typically a few hundred milliseconds end to end).

Authentication is identical to the Enrichment API. Send the same partner key in the X-API-Key header. A missing, invalid, inactive, or expired key returns 401.

The identifier

Send exactly one identifier as {"identifier": "..."}. We detect its kind automatically from its format, so you do not declare a type. For backward compatibility, the legacy {"phone": "..."} and {"sha256": "..."} fields are still accepted.

Response modes

Choose what comes back with the optional mode field. The default is enrich. Any other value returns 400. On a miss every mode returns {"matched": false}. Null fields are always omitted, so a result only carries the attributes we actually hold.

Returns the populated PII and demographic fields for the resolved person, the same data points the batch /api/v1/enrich endpoint returns (including phones, email, IP, and source URL). Only non-null fields come back, so the example below shows a representative subset; the full set is in the table.

curl -X POST https://www.talkdatatome.online/api/v1/resolve \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"identifier": "4155550142"}'

{
  "identifier": "4155550142"
}

{
  "matched": true,
  "first_name": "Jordan",
  "last_name": "Avery",
  "address": "1240 Oak Ridge Dr",
  "city": "San Francisco",
  "state": "CA",
  "zip": "94110",
  "age": 47,
  "income_band": "L. $150,000-$174,999",
  "income_midpoint": 162500,
  "credit_range": "A. 800+",
  "credit_midpoint": 820,
  "homeowner": "Home Owner",
  "email": "jordan.avery@example.com"
}

For real-time bidding. Send the identifier plus pre-qual filters and get back only {"matched", "pass"}, never any PII. Filters can sit at the top level or under a filters object. Every filter you send must pass, and a filter referencing a field the profile does not carry fails (we will not assert a qualification we have no data for). A miss, or any failed filter, returns "pass": false.

curl -X POST https://www.talkdatatome.online/api/v1/resolve \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{
    "identifier": "4155550142",
    "mode": "rtb",
    "state": "CA",
    "age_min": 35,
    "income_min": 100000,
    "homeowner": true
  }'

{
  "identifier": "4155550142",
  "mode": "rtb",
  "state": "CA",
  "age_min": 35,
  "income_min": 100000,
  "homeowner": true
}

{
  "matched": true,
  "pass": true
}

Pre-qual filters (all optional)

Returns only the identity keys for the resolved person (hashes, IP, source URL, and registration date), with no demographics. Useful for pixel resolution and identity stitching. Only non-null fields come back.

curl -X POST https://www.talkdatatome.online/api/v1/resolve \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{
    "identifier": "5d41402abc4b2a76b9719d911017c592",
    "mode": "identity"
  }'

{
  "identifier": "5d41402abc4b2a76b9719d911017c592",
  "mode": "identity"
}

{
  "matched": true,
  "md5": "5d41402abc4b2a76b9719d911017c592",
  "sha256": "9b2c…e4f1",
  "ip": "73.92.18.204",
  "source_url": "homeimprovementquotes.com",
  "registration_date": "2024-11-03"
}

Error codes

Errors return a JSON body shaped like {"error": "..."} with the matching HTTP status. Note there is no rate-cap status here: a miss is a normal 200 with "matched": false.

The reverse of the Enrichment API. Send up to 100 plaintext emails and/or up to 100 name-and-address records per call, and get back each person's phones (best number flagged), plus their name and postal address. Built for CRM phone-fill and outreach list preparation. Plan for 1 to 3 seconds per batch.

Authentication is identical to the other endpoints: send your partner key in the X-API-Key header. Emails must be plaintext (we hash them server-side to hit the index; a pre-hashed value cannot be converted, so it will not match). Records match through the postal graph first; when only the record's email matches, a last-name gate keeps us from returning a different person, and a same-surname, different-first-name match comes back flagged at low confidence. We log match counts only: the emails you submit and the phones we return are never persisted.

curl -X POST https://www.talkdatatome.online/api/v1/append-phone \
  -H "Content-Type: application/json" \
  -H "X-API-Key: YOUR_KEY" \
  -d '{"emails": ["jane.doe@example.com"]}'

{
  "records": [
    {
      "first_name": "Jane",
      "last_name": "Doe",
      "address": "123 Main St",
      "city": "Tampa",
      "state": "FL",
      "zip": "33601",
      "email": "jane.doe@example.com"
    }
  ]
}

{
  "requested": 1,
  "matched": 1,
  "results": [
    {
      "input": "jane.doe@example.com",
      "matched": true,
      "confidence": "medium",
      "first_name": "JANE",
      "last_name": "DOE",
      "address": "123 MAIN ST",
      "city": "TAMPA",
      "state": "FL",
      "zip": "33601",
      "phones": [
        { "phone": "8135550142", "type": "mobile", "rank": 1, "carrier": "T-MOBILE",
          "last_seen": "2026-04-18", "dnc": "callable", "best": true },
        { "phone": "8135550177", "type": "landline", "rank": 1, "carrier": null,
          "last_seen": "2025-11-02", "dnc": "callable", "best": false }
      ]
    }
  ]
}

Result fields

Email inputs come back in results (same order as submitted). Record inputs come back in record_results, each echoing the record you sent. A matched name and address also carries a demographics block (age, gender, income_band, credit_range, homeowner) from our own store; no id of any kind is needed. A miss is a normal 200 with "matched": false.

Each entry in phones

Error codes

The key you already hold works on everything below: your pixel data, property and vehicle lists, your feeds, delivery and spend reporting, your suppression list, live record intake, revenue reporting, and the RoofRadar endpoints. Each one is in the OpenAPI spec with full request and response schemas; this page gives you the call, the response, and the things worth knowing before you build on it.

Website visitors: your pixel data, with email and phone

The people your pixel resolved, as a file. GET counts them (free, nothing charged) and tells you how many carry an email, an address, a phone, and a cell phone, plus the top states and pages. Add any of the filters below as query parameters to size a selection. POST the same filters as JSON with a required max_records and get a secure download link to a CSV. Rows are never in the response body; a list belongs in a file you download.

The same export is one click in the app: open the Pixel menu, choose View your website funnel, then Export CSV. The API and the button produce the same file from the same selection.

curl "https://www.talkdatatome.online/api/v1/website-visitors?seen_within_days=30&require_phone=true" \
  -H "X-API-Key: rk_your_key_here"

{
  "available": true,
  "selection": "seen in the last 30 days, with a phone",
  "people": 23,
  "with_email": 23,
  "with_address": 23,
  "did_more_than_browse": 11,
  "with_phone": 23,
  "with_cell_phone": 17,
  "phone_no_dnc_flag": 17,
  "phone_household_only": 0,
  "top_states": [{ "state": "FL", "people": 4 }, { "state": "CA", "people": 3 }],
  "top_pages": [{ "page": "/", "people": 18 }, { "page": "/signup", "people": 9 }],
  "most_recent_visit": "2026-09-08",
  "total_people_in_database": 73,
  "notes": "people is distinct people, not visits. Free, nothing charged. POST the same filters with max_records to get a file.",
  "phone_notes": "phone_no_dnc_flag counts numbers we hold no do-not-call flag for, which is not the same as verified callable. ..."
}

curl -X POST https://www.talkdatatome.online/api/v1/website-visitors \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"seen_within_days": 30, "require_phone": true, "max_records": 5000}'

{
  "delivered": true,
  "rows": 23,
  "matched_before_cap": 23,
  "capped": false,
  "selection": "seen in the last 30 days, with a phone",
  "filename": "website_visitors_2026-09-08.csv",
  "download_url": "https://www.talkdatatome.online/api/downloads/...",
  "order_id": "2f1c0e4a-6b0e-4a1d-9c7e-3a2b1c0d9e8f",
  "note": "Secure link, valid 24 hours. One row per person."
}

Customers can export up to 200,000 people per call. The link is valid for 24 hours. When nobody matches, the response is {"delivered": false, "rows": 0} and nothing is charged. Before the pixel has resolved anyone, GET answers {"available": false} with a 200, not an error.

Filters (query parameters on GET, JSON fields on POST)

Columns in the file, in order

A phone never ships as digits alone. phone_type says whether it is a cell, phone_dnc whether it carries a do-not-call flag, and phone_match whether it is that person’s own line or their household’s, so you can decide what is dialable rather than reading the whole file as a call list.

Property pull: homeowners by value, year built, and equity

Homeowners in single-family homes, narrowed by an income floor, a home value window, a year-built window, and equity (the share of the home they own). Send estimate: true for a free funnel of counts; send max_records (1 to 100,000) instead to order the list, which is emailed, listed by GET /api/v1/pull, and downloadable below. Your saved suppression lists are already reflected in both.

# Free estimate
curl -X POST https://www.talkdatatome.online/api/v1/property-pull \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"estimate": true, "states": ["MI"], "min_home_value": 200000}'

# Order the list (same filters, max_records instead of estimate)
curl -X POST https://www.talkdatatome.online/api/v1/property-pull \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"states": ["MI"], "min_home_value": 200000, "min_equity_pct": 30, "max_records": 10000}'
# → 202 {"queued": true, "order_id": "...", "max_records": 10000, ...}

{
  "selection": "homeowners; single family; home value $200,000+; states: MI; one primary cell per person; DNC delivered as a flag column, not filtered",
  "people_in_geo": 9012568,
  "homeowner_base": 5690091,
  "with_qualifying_property": 1904376,
  "value_qualified": 1445245,
  "equity_qualified": 1445245,
  "unique_cells": 1130562,
  "dnc_flagged": 568995,
  "callable_non_dnc": 561567,
  "notes": "unique_cells = deliverable unique mobile numbers with DNC-flagged numbers INCLUDED (the file carries a DNC column). callable_non_dnc is the subset not on the DNC registry. Saved account suppression lists are already reflected."
}

Filters: states, zips (five digits or a three-digit prefix), cities, min_hhi, min_home_value, max_home_value, year_built_min, year_built_max, min_equity_pct, single_family_only, and cell_scope (primary, one cell per person, or all_cells). DNC is a flag, never a filter: numbers on the registry ship with a DNC column set to Y and are counted on both sides of the split, so your compliance decides. When an estimate cannot apply a filter it says so in ignored_filters with a warning, and the count reads high; do not quote it as if the filter had applied. An estimate is free of charge and never touches your balance or your monthly plan, but it is not unconditional: a key that is not linked to an account returns 403, and if your saved suppression lists cannot be read we return 503 rather than a count that quietly ignores them. Ordering needs list-pull access on the account.

The delivered file ends with Owner Status (Home Owner or Probable Home Owner) and Dwelling Type (Single Family and similar), so you can verify the homeowner and single-family selection on every row instead of taking it on trust. Both were appended after the existing columns, so nothing an importer already reads has moved.

Aged opt-in by vehicle: direct mail files

Aged opt-in auto insurance or auto warranty records (vertical), filtered by the vehicle on the form: vehicle_year_min and vehicle_year_max, vehicle_makes and vehicle_models to include or exclude, and exclude_heavy_duty, plus states, ZIPs, cities, and require_phone, require_email, require_vin. mode: "estimate" is free; mode: "order" with max_records (up to 200,000) queues the file, emailed and listed by GET /api/v1/pull.

curl -X POST https://www.talkdatatome.online/api/v1/aged-optin \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"mode": "estimate", "states": ["TX"], "vehicle_year_min": 2018, "exclude_heavy_duty": true}'
# Then order: {"mode": "order", ...same filters..., "max_records": 25000}  → 202

{
  "mode": "estimate",
  "selection": "auto insurance; states: TX; vehicle year 2018+; heavy duty excluded",
  "mailable_households": 48210,
  "with_phone": 41870,
  "with_email": 39955,
  "with_vin": 3104,
  "by_vehicle_year": [
    { "year": 2018, "households": 14120 },
    { "year": 2019, "households": 12904 }
  ],
  "notes": "mailable_households counts MAILABLE HOUSEHOLDS (name + street address + ZIP present, one row per household, newest form fill wins). This is the number that ships. ..."
}

Every count and every row is a mailable household: name, street address, and ZIP present, one row per household, newest form fill wins. That is smaller than a raw record count on purpose, because the count you are quoted has to be the count that ships. VIN coverage is low on this source, so quote with_vin before promising VINs. The file carries First_Name, Last_Name, Address, City, State, Zip, Vehicle_Make, Vehicle_Model, Vehicle_Year, and VIN, plus Phone and Email when include_contact is true; the phone is as supplied and not DNC-scrubbed, for record matching rather than dialing. For a second batch of the same selection, order again with offset set to the number already delivered.

Download a finished order

The CSV for any completed order on your account, byte-for-byte the file that was emailed. Orders from /api/v1/pull, /api/v1/property-pull, and /api/v1/aged-optin all download here; the id comes from GET /api/v1/pull, which lists recent orders with a download_url once the file exists.

curl -L https://www.talkdatatome.online/api/v1/pull/975a3e48-d46a-4afa-80ed-a544b69f4ff5/download \
  -H "X-API-Key: rk_your_key_here" -o list.csv
# 409 {"error": "Order is running; no file yet. Poll GET /api/v1/pull."} until the file exists

Topic search: what you can target

Roughly 50,000 targetable topics across seven universes: consumer and business in-market intent, keyword intent by funnel stage, inbound lead and caller verticals, live daily flows, and responder pools. Search with q, narrow with universe, category, or min_reach, page with limit and offset. The topic_id values are what POST /api/v1/pull accepts in topics. Metadata only: no PII, no records, nothing billed.

curl "https://www.talkdatatome.online/api/v1/topics/search?q=solar&limit=2" \
  -H "X-API-Key: rk_your_key_here"
# No parameters at all returns the universe summary

{
  "query": "solar",
  "total": 122,
  "limit": 2,
  "offset": 0,
  "results": [
    { "topic_id": "inbound_leads_solar", "universe": "inbound_leads", "label": "Solar", "category": "Inbound", "description": "Opt-in lead records in this vertical.", "reach": 3740260 },
    { "topic_id": "responders_solar", "universe": "responders", "label": "Solar", "category": "Inbound", "description": "Known responders in this vertical.", "reach": 5036 }
  ],
  "universe_labels": {
    "intent_b2c": "Consumer in-market intent topics",
    "intent_b2b": "Business in-market intent topics",
    "keyword": "Keyword search intent, by funnel stage",
    "inbound_leads": "Opt-in lead records by vertical",
    "inbound_calls": "Inbound callers by vertical",
    "inbound_live": "Live daily lead flow by vertical",
    "responders": "Known responders by vertical"
  }
}

reach is how many people are addressable for that topic and is omitted for live daily flows, where no fixed total exists. Only the two intent universes can be pulled on demand; the others are delivered as managed feeds.

Your feeds, described in plain English

Every feed on the account with the same description every other surface renders: what it delivers, what the records are made of (the measured mix once enough runs exist), every active filter, the cap, the schedule, how repeats are handled, and where it goes with credentials removed. recent_changes is the last ten changes, each saying who made it. Read-only and free. The id is what the two endpoints below take.

curl https://www.talkdatatome.online/api/v1/feeds -H "X-API-Key: rk_your_key_here"

{
  "count": 1,
  "feeds": [
    {
      "id": "996e114d-0fe5-414b-bb5d-9c075eb814b1",
      "name": "Final expense intent, daily",
      "active": true,
      "age_floor": null,
      "description": "Delivers audience people showing in-market web intent. 100% in-market web intent signals, no fill. ... Up to 5,000 records per run. Schedule: Daily at 7am ET. Repeat records are not filtered out. Delivered one record at a time to your webhook at ..., with your field mapping applied.",
      "characteristics": {
        "delivers": "Delivers audience people showing in-market web intent.",
        "composition": { "classes": [{ "class": "web_intent", "label": "in-market web intent signals" }], "single": true, "measured": null, "window": null, "text": "100% in-market web intent signals, no fill." },
        "filters": [],
        "cap": "Up to 5,000 records per run",
        "schedule": "Daily at 7am ET",
        "dedup": "Repeat records are not filtered out",
        "destination": "Delivered one record at a time to your webhook at ..., with your field mapping applied"
      },
      "recent_changes": [],
      "last_run_at": "2026-09-08T11:00:04.120000+00:00",
      "next_run_at": "2026-09-09T11:00:00+00:00",
      "manage_url": "/data-feeds/996e114d-0fe5-414b-bb5d-9c075eb814b1"
    }
  ]
}

Change a feed’s targeting: preview, then confirm

Two steps by contract. A call without confirm: true never changes anything: it returns the measured volume impact, from the feed’s own recent records, plus a confirm_token. Applying needs both confirm: true and that token, so a change cannot be one-shotted without somebody having seen the numbers. Tokens are good for about an hour; a stale one answers 409.

# Step 1: preview (nothing changes)
curl -X POST https://www.talkdatatome.online/api/v1/feeds/change \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"feed_id": "996e114d-0fe5-414b-bb5d-9c075eb814b1", "age_floor": 65}'

# Step 2: apply, with the token from the preview
curl -X POST https://www.talkdatatome.online/api/v1/feeds/change \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"feed_id": "996e114d-0fe5-414b-bb5d-9c075eb814b1", "age_floor": 65, "confirm": true, "confirm_token": "9f3a1c..."}'

// Preview
{
  "applied": false,
  "feed": "Final expense intent, daily",
  "impact": { "current_floor": null, "proposed_floor": 65, ... },
  "already_satisfied": false,
  "confirm_token": "9f3a1c...",
  "next_step": "Show these numbers to the person you are working for and get their explicit yes. Then call this endpoint again with the same change plus confirm: true and this confirm_token."
}

// Applied
{
  "applied": true,
  "verified": true,
  "feed": "Final expense intent, daily",
  "age_floor_before": null,
  "age_floor_after": 65,
  "takes_effect": "next run",
  "description": "Delivers audience people showing in-market web intent. ... Filters: Age 65 and up. ..."
}

Only the fields you can edit on the feed page move here. Today that is the age floor: age_floor (18 to 99) or remove_age_floor: true. Endpoints, credentials, and new fields stay person-only. An applied change is read back from the saved row before the response says verified: true, lands in the feed’s change history as made by you, and takes effect on the next run.

Pause, resume, or cap a feed

action is pause, resume, or set_daily_cap (with daily_cap). The same module the campaign chat uses, so pausing here means exactly what pausing there means. The account comes from the key, never from the body; a feed on another account answers 404.

curl -X POST https://www.talkdatatome.online/api/v1/feeds/control \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"feed_id": "996e114d-0fe5-414b-bb5d-9c075eb814b1", "action": "set_daily_cap", "daily_cap": 2500}'
# action: "pause" | "resume" | "set_daily_cap"

{
  "applied": true,
  "pending_approval": false,
  "kind": "data_feed",
  "feed_id": "996e114d-0fe5-414b-bb5d-9c075eb814b1",
  "daily_cap": 2500,
  "message": "Daily cap set to 2,500. Takes effect on the next run."
}

Going live is owner-gated. Resuming a paused data feed files an approval for the account owner instead of switching it on, and the response says so: "pending_approval": true, "applied": false. A daily cap above the plan ceiling is refused with 409 and nothing changes. A no-op (already paused, same cap) answers "applied": false with a message.

Audience delivery: what went out, per day

Per ad platform audience, per day, how many records were delivered. ?days sets the rolling window (1 to 365, default 30), ?audience_id limits to one audience, and ?include_daily=false drops the per_day series when you only need the summary. Every number is records net new to that audience on that date: a person already on the list is never sent again, so repeats are never counted. Free, read-only, and never PII. The same module feeds the Overview page.

curl "https://www.talkdatatome.online/api/v1/audience-delivery?days=30&include_daily=false" \
  -H "X-API-Key: rk_your_key_here"

{
  "window": { "days": 30, "from": "2026-08-10", "to": "2026-09-08", "today_is_partial": true },
  "audiences": [
    {
      "audience_id": "a74f0fc8288c9e6f",
      "name": "Homeowners 55+, Midwest",
      "platform": "Roku",
      "active": true,
      "sends_new_rows_daily": true,
      "cadence": "daily",
      "status": "delivering",
      "summary": "9,000 new records went out on Aug 16, then about 300 a day since.",
      "delivered_in_window": 15600,
      "days_in_window": 30,
      "days_with_delivery": 23,
      "average_per_delivery_day": 678,
      "average_per_calendar_day": 520,
      "best_day": { "date": "2026-08-16", "records": 9000 },
      "last_delivery": { "date": "2026-09-08", "records": 290 }
    }
  ],
  "totals": { "audiences": 1, "delivering": 1, "delivered_in_window": 15600 },
  "unavailable": false,
  "counts_what": "Every number is records NET NEW to that audience on that date. A person already on the list is not sent again, so repeats are never counted here."
}

Ad spend: the numbers your invoice is built from

What your connected ad accounts spent, from the same rows our billing reads, so an invoice and this endpoint cannot disagree. start and end are YYYY-MM-DD (default the last 30 days); group_by is day, account, campaign, or platform; and platform narrows to one. Dates are the ad account’s own reporting timezone, exactly as the platform reported them, so they match its dashboard across month boundaries.

curl "https://www.talkdatatome.online/api/v1/ad-spend?start=2026-08-01&end=2026-08-31&group_by=day" \
  -H "X-API-Key: rk_your_key_here"
# group_by: day | account | campaign | platform;  platform: meta | google_ads | vibe

{
  "start": "2026-08-01",
  "end": "2026-08-31",
  "group_by": "day",
  "total_spend": 4812.4,
  "currency": "USD",
  "all_settled": false,
  "days": 31,
  "results": [
    { "key": "2026-08-01", "date": "2026-08-01", "spend": 161.2, "locked": 161.2, "settled": true, "impressions": 48120, "clicks": 391, "currency": "USD" },
    { "key": "2026-08-31", "date": "2026-08-31", "spend": 149.9, "locked": null, "settled": false, "impressions": 45010, "clicks": 355, "currency": "USD" }
  ]
}

Two fields to read before reconciling

Suppressions: your do-not-contact list

Grow the account’s do-not-contact list from your own systems. POST phones (up to 50,000 per call, ten-digit US numbers, a leading 1 accepted) and they are excluded from every pull on the account from the next one on, and blocked on the realtime API too, with no other step. Additive only: nothing here ever removes a number, and repeats are ignored, so re-posting is safe. GET returns the counts, split by how the numbers arrived.

curl -X POST https://www.talkdatatome.online/api/v1/suppressions \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"phones": ["4155550142", "1 (212) 555-0188"]}'
# → {"ok": true, "received": 2, "added": 2, "total_phones": 214}

curl https://www.talkdatatome.online/api/v1/suppressions -H "X-API-Key: rk_your_key_here"
# → {"phones_uploaded": 12480, "phones_via_api": 214, "total": 12694}

Your database, in market: count who is shopping, then take the records

Upload your own database once, then ask who in it is in market right now and take those records, all from code. POST /api/v1/activation/datasets creates a database from a records array (up to 50,000 per call) or a CSV (as {"csv": "..."}, a text/csv body, or a multipart file, up to 1,000,000 rows), and matches every row to a person. Identity fields are detected by name; every other column is yours and is kept with the row, so an agency sends one file with a client column on every row. GET lists your databases with the columns each can be filtered on.

POST .../{id}/in-market answers "how many in my database are in market for roofing where client is acme". topic is free text resolved against the topic list (the response says what it matched), or name exact topic_ids. Leave the topic off for everyone in market for anything, with every topic each person touched on the row. where narrows to your own columns: {"client": "acme"} or a list of values. A column the database does not hold answers 422 and names the ones it does, so a misspelled tag is never a quiet zero.

# 1. Create the database once (your own columns such as client are kept with each row)
curl -X POST https://www.talkdatatome.online/api/v1/activation/datasets \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"name": "Q3 clients", "records": [
        {"email": "jane.doe@example.com", "phone": "4155550142", "first_name": "Jane", "last_name": "Doe", "zip": "94110", "client": "acme"},
        {"email": "sam@example.com", "zip": "10001", "client": "globex"}
      ]}'

# 2. Count who is in market for roofing where client is acme (free)
curl -X POST https://www.talkdatatome.online/api/v1/activation/datasets/<dataset_id>/in-market \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"topic": "roofing", "where": {"client": "acme"}, "recency_days": 7}'

# 3. Take those records (billed per record). Leave topic off for anyone in market for anything.
curl -X POST https://www.talkdatatome.online/api/v1/activation/datasets/<dataset_id>/in-market \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"mode": "pull", "topic": "roofing", "where": {"client": "acme"}, "recency_days": 7, "max_records": 5000}'

// Estimate
{
  "mode": "estimate",
  "database": "Q3 clients",
  "dataset_id": "0d3b6a1e-2c4f-4e8a-9b7d-5f6e7a8b9c0d",
  "rows": 48210,
  "matched_records": 41007,
  "topic": "roofing",
  "topics_matched": ["Roofing", "Roof Repair", "Roofing Contractors"],
  "recency_days": 7,
  "in_market": 312,
  "filtered_to": "client is acme",
  "breakdown": { "column": "client", "values": [{ "value": "acme", "in_market": 312 }] },
  "freshness": { "signal_date": "2026-09-22", "expected_lag_days": 2, "note": "In-market signals are a daily snapshot and normally land about 2 days behind. ..." },
  "next_step": "Pull them with mode \"pull\" and the same topic and where. 312 is the count that ships, before your account's caps."
}

// Pull
{
  "mode": "pull",
  "rows": 312,
  "columns": ["first_name", "last_name", "zip", "phone", "email", "current_mobile", "in_market_topics", "client", "List_ID"],
  "records": [
    { "first_name": "Jane", "last_name": "Doe", "zip": "94110", "phone": "4155550142", "email": "jane.doe@example.com",
      "current_mobile": "4155550142", "in_market_topics": "Roofing | Roof Repair", "client": "acme", "List_ID": "INMARKET-0D3B-20260924-A1B2C3D4" }
  ],
  "order_id": "a1b2c3d4-5e6f-7a8b-9c0d-1e2f3a4b5c6d",
  "list_id": "INMARKET-0D3B-20260924-A1B2C3D4",
  "download_url": "https://www.talkdatatome.online/api/v1/pull/a1b2c3d4-5e6f-7a8b-9c0d-1e2f3a4b5c6d/download",
  "charged_records": 312,
  "freshness": { "signal_date": "2026-09-22", "expected_lag_days": 2, "note": "..." }
}

mode: estimate (the default) is free and returns the count plus a breakdown by your column: the first column named in where, or any column you name in breakdown_by (leave where off and set "breakdown_by": "client" for every client at once). mode: pull returns the records: the identifiers you sent, current_mobile (the best mobile we hold), the topics each person is shopping, your columns, and List_ID on every row. It is billed per record through the same meter as any pull on the account and refused with a 402 up front when the plan or the prepaid balance cannot cover max_records. The file is also stored, so it lists under GET /api/v1/pull and downloads from GET /api/v1/pull/{order_id}/download; send format: csv to get the file straight back (the list id and the signal date then ride in the X-List-Id and X-Signal-Date headers). Nothing is charged for a count; a pull is charged only for the records it returns, so an empty result costs nothing. Every response carries freshness.signal_date: in-market signals are a daily snapshot and normally land about two days behind, so read a count as this week rather than this morning, and recency_days (1 to 30, default 3) counts back from that date. The same four steps are available to a connected assistant as the list_my_databases, create_my_database, estimate_my_database_in_market and pull_my_database_in_market tools, and signed in on the Database Activation page, step 3.

Live record intake for a connected database

Point a website form, a CRM automation, or a Zapier webhook at this URL and every new sign-up lands in the connected database as it happens, matched the same way the batch loaders match. The body is one record, an array, or {"records": [...]}, up to 500 per call. Field names are detected the way CSV headers are ("Email Address", "phone_number", "First Name", zip or postal). Each record needs an email or a phone. Repeats and webhook retries are deduplicated, never double-inserted. The dataset id is on the Database Activation page.

# Dry run first: checks the key, the database, and the field names, stores nothing
curl -X POST "https://www.talkdatatome.online/api/v1/activation/datasets/<dataset_id>/records?dry_run=1" \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"Email Address": "jane.doe@example.com", "phone_number": "4155550142", "First Name": "Jane"}'

# Then the real call: one record, an array, or {"records": [...]}, up to 500 per call
curl -X POST https://www.talkdatatome.online/api/v1/activation/datasets/<dataset_id>/records \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"records": [{"email": "jane.doe@example.com", "zip": "94110"}, {"phone": "2125550188"}]}'

// Dry run
{
  "ok": true,
  "dry_run": true,
  "dataset_id": "0d3b6a1e-2c4f-4e8a-9b7d-5f6e7a8b9c0d",
  "dataset_name": "Newsletter sign-ups",
  "received": 1,
  "would_insert": 1,
  "unusable": 0,
  "records": [{ "index": 0, "usable": true, "matched_on": "email", "read": { "email": "jane.doe@example.com", "phone": "4155550142", "first_name": "Jane" }, "ignored": [], "dropped": [] }],
  "note": "Your key and this database check out. Nothing was stored. Send the same call without dry_run to load it for real."
}

// Live
{
  "ok": true,
  "dataset_id": "0d3b6a1e-2c4f-4e8a-9b7d-5f6e7a8b9c0d",
  "received": 2,
  "inserted": 2,
  "matched": 1,
  "duplicates": 0,
  "invalid": 0,
  "dataset": { "rows": 1204, "matched": 871 }
}

Test mode: add ?dry_run=1 (or "dry_run": true in the body) and we check the key, the database, and every field name we can read out of your payload, then store nothing. It is the same code path right up to the insert, so a green dry run means the real call will work; the Test button on the Database Activation page is this call. Form tools that cannot set headers may pass the key as ?key= instead. An unknown dataset answers 404.

Revenue reports: what a delivered list earned

For buyers on a revenue share. Every list we deliver carries a list_id on the file or in the payload, like FINALEXPEN-996E-20260824-222CFF7A. Post what you owe TDTM against it as a total for a period (reports) or one sale at a time (transactions). Both price identically; sales are better evidence. Up to 500 items per call. GET reads back what you reported, and ?mode=lists returns the recent deliveries you can report against.

# A total for one list
curl -X POST https://www.talkdatatome.online/api/v1/revenue \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"reports": [{"list_id": "FINALEXPEN-996E-20260824-222CFF7A", "revenue": 4120.00}]}'

# Or every sale, one at a time
curl -X POST https://www.talkdatatome.online/api/v1/revenue \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"transactions": [{"list_id": "FINALEXPEN-996E-20260824-222CFF7A", "revenue": 62.50, "occurred_on": "2026-09-14", "record_ref": "lead-88"}]}'

# Name the person who converted and the sale ties to the record we sent.
# Plaintext or hashed, optional, and it never changes what you are paid.
curl -X POST https://www.talkdatatome.online/api/v1/revenue \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"transactions": [{"list_id": "FINALEXPEN-996E-20260824-222CFF7A", "revenue": 62.50, "occurred_on": "2026-09-14", "email": "jane.doe@example.com"}]}'

# Lost the file? The list ids you can report against:
curl "https://www.talkdatatome.online/api/v1/revenue?mode=lists" -H "X-API-Key: rk_your_key_here"

{
  "ok": true,
  "accepted": 1,
  "rejected": 0,
  "matched_to_a_delivery": 1,
  "transactions": 0,
  "period_totals": 1,
  "results": [
    {
      "index": 0,
      "status": "stored",
      "id": "7c2d...",
      "list_id": "FINALEXPEN-996E-20260824-222CFF7A",
      "feed_code": "FINALEXPEN-996E",
      "kind": "period",
      "occurred_on": null,
      "match": "matched_run",
      "revenue_cents": 412000,
      "currency": "USD",
      "revision": 1,
      "replaced_revenue_cents": null,
      "warnings": []
    }
  ]
}

// A sale that named the person who converted carries identity_match, which is
// honest about what was found: matched, matched_other_delivery, no_match, or
// not_checked. It never affects the money.
{
  "ok": true,
  "accepted": 1,
  "transactions": 1,
  "sales_tied_to_a_record": 1,
  "results": [
    { "index": 0, "status": "stored", "match": "matched_run", "identity_match": "matched", "revenue_cents": 6250 }
  ]
}

// GET ?mode=lists
{
  "ok": true,
  "lists": [
    { "list_id": "FINALEXPEN-996E-20260824-222CFF7A", "feed_code": "FINALEXPEN-996E", "feed": "Final expense intent, daily", "delivered_at": "2026-08-24T23:11:17.827542+00:00", "records": 10, "status": "completed" }
  ]
}

Three rules. Your figure is never lost: a report we cannot tie to a delivery is stored with "match": "unmatched" and returned with an explanation, not answered with a 400. Nothing is invented: unmatched money is shown for exactly what it is. Re-posting is safe: a report is keyed on what it is about (list, period, your own external_id), so a second post updates the row, bumps revision, and keeps the figure it replaced. Send both sales and the total they add up to and you are not counted twice; the sales win. Rows that fail validation are answered per row in results (a missing list_id, for example); the batch as a whole is 400 only when nothing in it could be stored.

Fields on each item

RoofRadar: lists and lookups for your territory

Three endpoints for RoofRadar accounts, on the same key. All three need a RoofRadar plan on the account and answer 402 without one. ZIPs have to be inside the plan’s territory (403 otherwise, with the ZIPs that were outside listed). Records count against the plan on the first download of a list, or on a matched lookup; a miss is free.

POST ZIPs (up to 1,000), a required max_records (1 to 100,000), and optional filters: built_on_or_before, built_on_or_after, min_home_value, max_home_value, min_income, min_owner_age, max_owner_age, phone (mobile, any, or none), and require_email. mode is homeowners (default) or intent (an add-on); columns is full or dialer. The list builds in the background and is emailed; GET lists your recent lists with download links and what is left on the plan.

curl -X POST https://www.talkdatatome.online/api/v1/roofradar/pull \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"zips": ["49503", "49504"], "max_records": 500, "filters": {"built_on_or_before": 2005, "phone": "mobile"}}'
# → 202 {"queued": true, "mode": "homeowners", "max_records": 500, "zips": 2, "warnings": [], "note": "Building now. ..."}

curl https://www.talkdatatome.online/api/v1/roofradar/pull -H "X-API-Key: rk_your_key_here"

{
  "lists": [
    {
      "id": "4b9eee20-5d9c-45d7-a8a3-308be83c7905",
      "created_at": "2026-09-08T14:02:11.000000+00:00",
      "completed_at": "2026-09-08T14:03:40.000000+00:00",
      "status": "completed",
      "records": 500,
      "mode": "homeowners",
      "summary": "built 2005 or earlier; mobile phone required",
      "downloaded": false,
      "download_url": "/api/v1/roofradar/lists/4b9eee20-5d9c-45d7-a8a3-308be83c7905/download",
      "ad_platform_url": null,
      "emailed_only": false,
      "error": null
    }
  ],
  "records_remaining": 9500,
  "period_end": "2026-10-01T00:00:00+00:00"
}

One home, its roofing facts, its owner. Send {"address", "zip"} or {"phone"}. A match costs one record; up to 60 lookups a minute per key. roof_cover is a label only when the code is known, otherwise null with the raw code in roof_cover_code, never a guess. Phones on your do-not-contact list are withheld and suppressed_contact is set.

curl -X POST https://www.talkdatatome.online/api/v1/roofradar/resolve \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"address": "745 Monroe Ave NW", "zip": "49503"}'
# or  -d '{"phone": "6165550100"}'

{
  "matched": true,
  "matched_by": "address",
  "billed": true,
  "records_remaining": 9499,
  "address": "745 MONROE AVE NW",
  "city": "GRAND RAPIDS",
  "state": "MI",
  "zip": "49503",
  "county": "KENT",
  "owner_occupied": true,
  "year_built": 1998,
  "roof_age_years": 28,
  "roof_cover": "Asphalt shingle",
  "roof_cover_code": 3,
  "living_sqft": 1840,
  "home_value": 312000,
  "owner": { "first_name": "Jordan", "last_name": "Avery", "mobile_phone": "6165550100", "landline_phone": null, "email": "jordan.avery@example.com", "age": 51, "household_income": 92500 }
}

The CSV for a finished list. The first download meters the records against the plan; later downloads are free. When the remaining allotment is short the download is refused with 402 rather than partially billed. ?format=hashed serves the ad platform file (SHA256 email and phone), only after the list itself has been downloaded once.

curl -L https://www.talkdatatome.online/api/v1/roofradar/lists/<list_id>/download \
  -H "X-API-Key: rk_your_key_here" -o roofing_list.csv
# Add ?format=hashed for the ad platform file, once the list itself has been downloaded

Identify (deprecated)

An alias of POST /api/v1/enrich with records: the same name-gated waterfall over phone, email, and name plus address, returning the identical shape. It stays only so integrations built on it keep working. New integrations should call /api/v1/enrich (documented above); for a single real-time phone, use /api/v1/resolve with verify.

# Deprecated. Send the same body to /api/v1/enrich instead:
curl -X POST https://www.talkdatatome.online/api/v1/enrich \
  -H "X-API-Key: rk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"records": [{"phone": "4045550142", "first_name": "Jane", "last_name": "Doe", "zip": "30301"}]}'

Status codes across these endpoints

Errors return a JSON body shaped like {"error": "..."} with the matching HTTP status, and the message says what to change. These are the ones you will meet.