Company profile · 4 min read
Crustdata: Live people and company data for agents
Crustdata provides live people and company context through APIs and MCP for sales, recruiting, investing and internal tools.
Cofounder of Lazyweb (a16z Speedrun). Previously grew subscriptions and built native ads at Duolingo. Writes First1000 (90,000+ subscribers).
Published · Updated
Crustdata is infrastructure for agents that need live people and company context, not a static lead list.
What it does
Crustdata offers company and people data through APIs and MCP. Its public product surface includes company and people search, enrichment, jobs, posts, watchers and a web-search API. The stated use cases are practical: AI SDRs, recruiting platforms, investment and diligence workflows, internal sales tools and monitoring company signals.
The web-search product is the clearest agent fit. Its API page describes a call, live search and fetch, then structured JSON returned for an agent to use. That is a different product from handing an agent a spreadsheet and hoping it can infer what is current.
Why I’d look closer
The company has an explicit data-layer position. Its homepage says the graph resolves people and companies from web sources and is available through API or MCP. The public pages also show the product split between search, enrichment and watching, which matters because different buyers need different kinds of freshness.
The pricing model reflects that distinction. Crustdata’s pricing page describes credit-based real-time API access, monthly or annual plans, developer support and a separate flat-file option refreshed monthly from more than 11 sources. Enterprise pricing is custom. There is no simple per-seat number to copy into a budget; the decision depends on whether you need live signals, bulk data or both.
The YC profile gives useful professional context without pretending it is a review. It describes Abhilash Chowdhary’s background in large-scale data pipelines and robotics, Manmohit Grewal’s previous data products, and Chris Pisarski’s work on a data platform. It also lists a 30-person team on the retained profile snapshot. Those details support the data-infrastructure thesis; they do not establish data accuracy.
What could make it the wrong choice
The obvious buyer questions are provenance, freshness and credit consumption. A live signal is only valuable if the source is allowed to be used, the record is current enough for the decision, and the API cost fits the workflow. MCP availability may simplify agent integration, but it does not remove the need to check fields, permissions and failure behavior.
The public source packet contains company-selected testimonials and use cases, not independent reviews. I would ask for sample records, refresh rules, deletion handling and a task-level accuracy definition before building a critical workflow on top of it.
My editorial take
I would shortlist Crustdata for an agent that needs changing company, people or market context: sales research, recruiting, investing or monitoring. It is a poor fit for a one-off list that does not need freshness. The product decision is whether live context changes the outcome enough to justify API complexity and credit-based spend.
Quick facts
| Field | Sourced detail |
|---|---|
| Product | Live people and company data through API or MCP |
| Buyers | Sales, recruiting, investing and internal data teams |
| Access model | Credit-based API plus monthly refreshed flat-file data |
| Public product surfaces | Search, enrichment, jobs, posts, watchers and web search |
| Pricing | Custom enterprise options; no fixed public numeric plan used here |
Sources checked
| Source | Checked |
|---|---|
| YC profile | 2026-09-19 |
| Crustdata homepage | 2026-09-19 |
| Pricing | 2026-09-19 |
| Web Search API | 2026-09-19 |
Cohort context
Crustdata is listed in Fall 2024. In our 2026-09-18 directory snapshot, 57 of 94 listed companies in that cohort have YC’s primary industry label B2B (60.6%). This is a current-directory comparison, not an original intake count or a performance ranking. Nine-cohort dataset.
Public website snapshot
Observed 2026-09-19T16:14:34.023Z in raw homepage HTML. This records visible metadata and advertised links, not agent execution or product quality.
| Signal | Homepage observation |
|---|---|
| Product description metadata | Observed |
| Canonical link | Observed |
| H1 or H2 heading | Observed |
| Typed structured data | Observed |
| Docs/developer link | Observed |
| Pricing link | Observed |
| llms.txt link | Not observed in this response |
| Markdown alternate | Not observed in this response |
Public observations · Collection method. Missing links here do not establish that a capability or file is absent elsewhere.
About the author
Cofounder of Lazyweb (a16z Speedrun). Previously grew subscriptions and built native ads at Duolingo. Writes First1000 (90,000+ subscribers).
I cofound Lazyweb and publish Mudpie. This is an owner-written publication, not an independent testing organization. Research notes distinguish observations, sourced reporting and editorial judgment.
First1000 ↗ · X ↗