Hey Product Hunt! 👋 We built Oliver because we kept running into a problem as AI agents started doing more real work: the work gets done, but the context behind it gets lost. A decision might happen in Slack, the implementation in GitHub, supporting context in Notion, and part of the reasoning inside an agent run. A few weeks later, neither a person nor another agent has the full picture of why something was done. Oliver connects that history. It captures what people and agents are doing, links it back to the underlying evidence, and makes that context searchable and reusable by future agents. The goal is that before an agent changes something, it can understand what happened before, why previous decisions were made, and avoid repeating work or undoing something intentionally. We're opening Oliver up for free with rate limits while we learn where this is most useful. We're especially looking for feedback from teams already using AI agents heavily.

Olivergraph

System of record for team and agent work

System of record for team and agent work