Context.ai
The business, the turning points, and the decisions behind them.
Some sections are still being researched1 / 4
Sections to complete：02
01 / WHAT IT BUILT
What the company did
Context.ai was a London startup founded in 2023 by former Google employees Henry Scott-Green (CEO) and Alex Gamble (CTO), building product analytics for LLM-powered applications. Recognizing that traditional analytics tools 'are built to count clicks' while text becomes the primary user interaction, it ingested user conversation transcripts and applied topic clustering, sentiment analysis, and feedback ratings so businesses could track frequently discussed topics, debug bad conversations, monitor brand risks, understand retention, and measure the impact of new releases. Customers included Cognosys, Lenny's Newsletter, Juicebox, and ChartGPT.
02 / FAILURE REASONS
Material for this section is being gathered and checked. It will appear here once ready.
03 / EVENT HISTORY
Events in the structured record
- Founded; $3.5M seed
Context.ai launched with $3.5 million in seed funding co-led by GV (Google Ventures) and Tomasz Tunguz at Theory Ventures, with participation from 20SALES and angels including Harry Stebbings, Guy Podjarny, and founders of Synthesia, DeepMind product leadership, GoCardless, deepset, Relay, and Predibase.
- OpenAI acqui-hire; products wound down
In April 2025, Context.ai announced that its co-founders would join OpenAI to work on model evaluations (with Scott-Green's LinkedIn listing him as a product manager 'building evals'), and that the startup would wind down its existing products. Financial terms were undisclosed.
04 / CONTEMPORARY REBUILD ASSESSMENT
Difficulty of rebuilding this project today
Material for this section is being gathered and checked. It will appear here once ready.
Overall assessment: Not completed
05 / CAUSAL CHAIN
How the factors connect
- 01
Raised approximately $3.56 million in total (reports cite a $3.5 million seed round).
- 02
In April 2025 its co-founders were acqui-hired by OpenAI to work on model evaluation, and the products were wound down; terms were not disclosed.
- 03
No audited loss figure was publicly disclosed.