AI training and web data companies: 10 verified, updated November 13, 2026
Presspool has verified 10 AI training and web data companies. 7 of them announced a round in the last six months, most recently Mecka ($60 million, October 7, 2026). Latest rounds: the 10 announced in dollars total $1.29 billion.
A subset of Presspool's AI infrastructure map: companies whose fit on that map places them in AI training and web data. Companies selling the layers other AI products run on: GPU clouds and neoclouds, inference and model-serving platforms, training data and web data for AI, evaluation, observability and testing for AI, vector and AI-native databases, and infrastructure for AI agents. Each company is already on the parent map, where it was taken from the company's own website and confirmed by an independent source, both linked on the entry. No company pays to be listed.
This map does not have its own weekly question set or answer sample. The questions and answer counts for these companies are on the AI infrastructure map.
| Company | What it does | Latest announced financing |
|---|---|---|
| Mercor | Mercor organizes human experts to power the AI economy, offering expert work for AI labs, custom evals and the APEX benchmarks of frontier-model performance. | $350M, Series C (October 27, 2025) |
| Snorkel AI | Snorkel AI builds expert-curated datasets and environments that AI labs, enterprises and government agencies use to train and evaluate models. | $350 million, Series E (September 22, 2026) |
| Exa | Exa offers search over a large index of public web and private information, delivered through APIs built for AI agents. | $250M, Series C (May 20, 2026) |
| Parallel Web Systems | Parallel builds web infrastructure that lets AI systems search, extract, monitor and reason over information on the web. | $100M, Series B (April 29, 2026) |
| Firecrawl | Firecrawl is a web data API that searches, scrapes and interacts with websites, turning them into clean Markdown or structured data for AI agents. | $75 million, Series B (September 22, 2026) |
| Encord | Encord builds data infrastructure for AI teams to manage, curate and annotate multimodal training data, with a focus on physical AI. | $60 million, Series C (February 26, 2026) |
| Mecka | Mecka is the data and deployment layer for physical AI, turning real-world activity into datasets robotics and AI teams train on. | $60 million, Series B (October 7, 2026) |
| Halluminate | Halluminate builds benchmarks and RL environments that train frontier AI models on financial-services work. | $30 million, Series A (October 1, 2026) |
| HUD | HUD is a platform for building reinforcement-learning environments that encode domain expertise to train and evaluate models and produce post-training data. | $16M, Series A (June 30, 2026) |
| Ocular AI | Ocular AI provides research, evals and expert data for AI. | $2 million, Pre-seed (October 5, 2026) |
AI training and web data startups that raised funding recently
Rounds announced in the six months to November 12, 2026, newest first. Each links the company's record and its sources.
- Mecka: $60 million, Series B, announced October 7, 2026, led by Sequoia Capital.
- Ocular AI: $2 million, Pre-seed, announced October 5, 2026, led by Drive Capital.
- Halluminate: $30 million, Series A, announced October 1, 2026, led by Oak HC/FT.
- Firecrawl: $75 million, Series B, announced September 22, 2026, led by Smash Capital.
- Snorkel AI: $350 million, Series E, announced September 22, 2026, led by Insight Partners and S32.
- HUD: $16M, Series A, announced June 30, 2026, led by Standard Capital.
- Exa: $250M, Series C, announced May 20, 2026, led by Andreessen Horowitz.
What counts as AI training and web data
This page lists companies already on the AI infrastructure map. A company is listed here when the fit recorded on that map names AI training and web data before any other layer. It is not a separate check. Out of scope: companies on the AI infrastructure map whose fit names a different layer first, and anything the AI infrastructure map excludes. Companies whose product is an end-user AI application, and companies whose most recent announced round is more than 12 months old. Definitions are versioned; this is version 1. Corrections need a public source: see Contact.
Changes
- November 13, 2026: Opened as a subset of the AI infrastructure map. Members are companies already on that map whose fit names this layer first. No separate company check was made for this page.