Hybrid retrieval
BM25 plus dense embeddings fused with reciprocal rank fusion, with every citation validated against the source page.
AI agent · climate risk
Projects Climate-Risk Agent
Ask about heat, rain or wind risk for any place on Earth. It answers with a report that cites its sources, or refuses when the evidence is not there.
Try the live app ↗
{"answer": "","citations": [],"abstain": true,"abstain_reason": "The provided excerpts do not contain information specific to Rourkela or short-term forecasts (next 7 days) for heatwaves."}A question about heat, extreme rain or wind at one place needs three kinds of evidence at once: a live forecast, the local record of extremes from ERA5, and what IPCC AR6 assesses for that region.
Both eval sets are SHA-256 frozen. The 105-question held-out set was written after the dev set existed and never used to tune anything.
Refusals are scored with the answers: correct answer, correct refusal, false refusal, false answer. Any false answer fails the build.
The agent also measures its own forecast skill per lead day and weights the confidence of every report by it.
held-out retrieval run, 2026-09-07 (JSON)evaluation section of the README
The question is parsed, geocoded and mapped to its IPCC AR6 region before the agent sees it. The agent can refuse at its first step; otherwise it fetches the forecast, searches AR6, reads the regional projection and writes the report. The flow follows the diagram in the repository README.
Before the agent
The agent · LangGraph
Feeds call and synthesize
ERA5 to GEV hazard statisticsstationary and drifting-location GEV · likelihood-ratio trend test · return levels with 90% bootstrap CIFeeds research and synthesize
IPCC AR6 retrievalBM25 + dense hybrid, RRF fusion · citations validated against the retrieved pagesBM25 plus dense embeddings fused with reciprocal rank fusion, with every citation validated against the source page.
60+ years of ERA5 fitted with stationary and non-stationary GEV distributions, a likelihood-ratio trend test, and 90% bootstrap confidence intervals.
Two MCP servers ship with it, weather and ipcc-rag, on protocol revision 2026-07-28 with ToolAnnotations. The IPCC server is indexed on the official MCP registry; its container image is private for now, so run it from a clone.
Records per-request telemetry and cost, with a Cloud Storage cache in production (Redis optional), a disk fallback and a prewarm script for the hazard fits.
READMEload_dotenv(override=False).1,138 tests passed in the CI run on main · 7 skippedCI run · Nov 2026
Contact / Rourkela, Odisha, India