Collect, Store, Analyze, and Act
About HypAware
Collect, store, analyze, and act on your AI logs
Agents, coding tools, and chatbots generate quadrillions of tokens of conversational and tool-call data every year. Buried in those logs is everything teams need to understand what their AI is actually doing in production: where conversations go off the rails, where tool calls fail, which prompts burn tokens, and how each release shifts behavior.
But existing tools are not built for this. Observability dashboards aggregate, sample, and flatten. Jupyter notebooks and warehouse SQL choke on multi-gigabyte JSONL of nested conversations. No one can read through millions of rows of text, and the most important signal often lives in a long-tail 1% you would never spot with sampling.
Built on Hyperparam
Hyperparam Corporation in Seattle built Hyperparam, a tool for exploring and transforming AI logs at a scale nothing else could read. Its engine, client-native Parquet and Iceberg readers with grep, SQL, and graph on top, is what makes HypAware possible. HypAware is the next product on that foundation, pointed at one job.
HypAware is the improvement loop for AI agents: collection, storage, analysis, and action. A free, open source CLI hooks into Claude Code, Codex, Cursor, OpenClaw, Hermes, and anything else that speaks OpenTelemetry, and writes every session as Apache Iceberg tables. HypAware reads those sessions at scale, finds what is broken, hands your agent the fix, and checks on the next pass that it held. HypAware Cloud runs that loop across a whole team and its fleet.
Hyperparam is the exploration and transformation tool. It reads agent traces, coding-tool transcripts, and chatbot histories straight from where they land and pairs them with an AI agent that analyzes the logs alongside you: drill into nested traces, generate derived columns at scale, build SQL views that join across sources, and save reusable analyses as skills.
Open formats, open code
Sessions are stored as Apache Iceberg tables, an open format any engine can read. The collector is open source and can be read before it is run. The path from your agents to your findings holds no proprietary format and no warehouse; the parts we keep closed are in HypAware Cloud, not in your data.
Open source
The libraries underneath both products, hyparquet, icebird, hightable, and squirreling, and the HypAware CLI itself, are at github.com/hyparam.
The company
HypAware is built by Hyperparam Corporation, the company behind hyperparam.app. The founder, team, and investors are on hyperparam.app/about.