19 buyers of business records for AI training: what each takes, who it accepts and how it pays.
Checked 19 October 2026. Figures are each buyer's own published ranges, not offers. TydeX is independent and is not affiliated with any buyer listed. How we research
19 buyers
| Buyer | Type | What it takes | Who it accepts | Published range | How it pays |
|---|---|---|---|---|---|
| Appen | Data lab | Operational documentation, Decision and escalation records, Systems of record such as ERP, contracts and procurement | No published criteria · Not stated | Not published | on acceptance |
| Centific | Data lab | Approval chains and exception handling, Workflows that run across several business applications, Operational records from 188 supported apps | No published minimum · Not stated | Not published | on acceptance |
| DataVendor | Marketplace | Private codebases | Staff minimum not stated · Not stated | Not published | after sale |
| Grepped | Marketplace | Messages, documents and tickets, Code repositories, CRM, accounting and support records | No published minimum · Not stated | $20k to $5m | after sale |
| Handshake AI | Data lab | Email, documents and chat, Design files such as AutoCAD and Figma, Business software such as QuickBooks, NetSuite, Notion and ServiceTitan | 20+ full-time employees · Not stated | $100k to $4m+ | Not stated |
| Labelbox Prism | Data lab | Work documents, Operating data from Salesforce, Jira, SAP and CAD, Financial and scientific reasoning records | No published minimum · Not stated | $50k to $1m+ | on signing, revenue share |
| Mercor | Data lab | Workflow data from Slack, Teams, Gmail and Outlook, Salesforce, Confluence, Drive and Notion records, GitHub repositories, commits, pull requests and issues | No minimum published · Not stated | Not published | on delivery |
| micro1 | Data lab | SOPs, knowledge bases and internal documentation, CRM data and project histories, QA processes and how teams make operational decisions | 30+ employees, with established documentation · Named | $100k+ | Not stated |
| Mode | Direct buyer | Email, documents and chat, CRM and billing records | 6+ for legal, 10+ for accounting, finance and medical, 30+ for others · US focus | $100k to $5m | Not stated |
| Nyne | Direct buyer | Chat, email and documents, CRM, HR, ERP and support records | 20+ full-time US W-2 employees · US focus | Not published | on acceptance |
| Ooak Data | Data lab | Documents, wikis, Jira, Notion and Trello, CRM, sales and financial records, Email, chat, tickets and code repositories | No published minimum. One example dataset is from a firm with 15+ people · Not stated | $20k+ | Not stated |
| Pewter | Broker | Email, files and business records | 20+ full-time staff, now or at peak · US only | $350k to $650k | Not stated |
| Replay | Direct buyer | Code and commits, Decision threads and support history, Financial, CRM and sales records | Price bands start at 20 employees. Headcount is a proxy · US focus | $10k+ | upfront, revenue share, recurring |
| Scale AI | Data lab | Records from data warehouses such as BigQuery, Snowflake and Databricks, Chat, email, calendars and documents, Tickets, CRM, HR and ERP records | 40+ employees preferred, assessed case by case · Not stated | $10k to $1m+ | recurring |
| SimpleClosure AssetHub | Wind-down specialist | Source code, Workspace data | Staff minimum not stated · Not stated | $10k to $100k | Not stated |
| SuperAnnotate | Data lab | Slack threads, email chains and approvals, Runbooks, wikis, tickets and incident reviews, Specialist records such as inspection reports and claims files | Any size · Anywhere | $100k to $1m+ | on delivery, recurring |
| Telegraph Lab | Direct buyer | Work conversations, Business software records, Company drives | Says it works across industries and company sizes · Not stated | $100k to $4m | Not stated |
| Troveo | Marketplace | Company data exports | Staff minimum not stated · Not stated | Not published | revenue share, after sale |
| Turing (Project Lazarus) | Data lab | Startup codebases, Operational data | Staff minimum not stated · Not stated | Not published | Not stated |
Missing a buyer? Email charlie@tydex.ai
