Post-Departure
Knowledge Recovery
When key employees leave, their institutional knowledge dies in email archives. PikoClaw resurrects it — surfacing contacts, decisions, and context from PST files in minutes. The Enron corpus is the proof point.
Start DemoUpload
Feed PikoClaw an email archive
Drop in a PST, MBOX, Maildir, or EML file. PikoClaw ingests any email format and begins extracting institutional knowledge in seconds.
Go to UploadExplore
Browse extracted knowledge
Automatically surfaces contacts, topics, and relationships from the archive. See the numbers: 23 contacts, 15 topics, 38 graph nodes — all structured and ready.
Go to ExploreSearch
Ask questions about the archive
Full-text and semantic search across every message. Try: "Q3 budget figures" — PikoClaw returns the answer with source citations.
Go to SearchAgent Pipeline
Watch multi-agent orchestration
Geryon (OpenClaw) coordinates PikoClaw's extraction, graph, and search workers in parallel — delivering structured intelligence faster than any single agent.
Go to AgentsChat
Conversational knowledge recovery
Ask open-ended questions in plain language. Chat synthesizes answers from email threads, contact graphs, and extracted topics — with full provenance.
Go to ChatBuilt With
Python
Extraction · PII redaction · Clustering
CF Workers / R2 / D1
Serverless API · Object storage · SQL
Next.js 14
App Router · RSC · Streaming
D3.js
Force-directed knowledge graph
OpenClaw
Multi-agent orchestration (Geryon)