# Relay > Relay is an organisational continuity platform. It records who is accountable for what (people and teams, primary owners and backups, any kind of work), rebuilds the context behind that work from GitHub, Linear and Notion with cited evidence, forecasts what an absence would break, and moves ownership through reviewed, audited handovers. The database is the source of truth; AI only proposes evidence-backed drafts and answers, and never changes ownership, approvals or source records. The live product is at https://relay-continuity-nine.vercel.app. It is a working multi-tenant SaaS: real WorkOS sign-in, real GitHub, Linear, Notion and Slack connections, and Postgres with row-level security. A pre-seeded sample organisation (and one shared account per role for judges) exists only so evaluators can see every feature and edge case without connecting their own tools; a workspace connected to real tools uses exactly the same code paths. Source code and documentation: https://github.com/prasadbhalerao1/CodersInPune ## Product - [Judge guide](https://github.com/prasadbhalerao1/CodersInPune/blob/main/docs/JUDGE_GUIDE.md): the problem, the continuity lifecycle, an implementation-evidence matrix and the judging criteria - [README and ten-minute walkthrough](https://github.com/prasadbhalerao1/CodersInPune/blob/main/README.md): what to click on the live product - [Project overview](https://github.com/prasadbhalerao1/CodersInPune/blob/main/docs/PROJECT_OVERVIEW.md): capabilities, architecture, AI evaluation, operations - [Competitive analysis](https://github.com/prasadbhalerao1/CodersInPune/blob/main/docs/competitive-analysis.md): service catalogues, knowledge tools, offboarding and incident tools compared, with sources - [Product guide](https://relay-continuity-nine.vercel.app/docs): every feature and every number, with its exact formula and the code that computes it; full text at [/llms-full.txt](https://relay-continuity-nine.vercel.app/llms-full.txt) - [Submission pack](https://github.com/prasadbhalerao1/CodersInPune/blob/main/docs/submission/README.md): workflow traceability, feature inventory, quality scorecard - [Round 3 improvements delivered](https://github.com/prasadbhalerao1/CodersInPune/blob/main/docs/round3/improvements-delivered.md): AI Handover Copilot, What If? simulator and Project Memory Replay, mapped to code and tests ## Core capabilities - Responsibilities with primary owners, co-owners and backups; ownership history; review cadences - Continuity risks from explainable rules (no owner, no backup, single holder, stale knowledge) and key-person exposure - Project Memory Replay: merged, cited history per responsibility; AI chapters labelled Inferred; cited Q&A - What If? simulator: deterministic, read-only impact of an absence, cover candidates with overload flags - Handover Copilot: per-recipient brief with Fact / Rule / AI inference labels, brief coverage, version comparison, approval and acknowledgement, cited Q&A that answers "Not in the evidence" rather than guessing - Handover Autopilot: one click runs the whole case analysis while you watch each tool call; ends in proposals a person accepts - Meeting intelligence: transcript in; decisions, actions, open questions and risks out, each citing its line; accepted items become decision records and case tasks - Ask Relay: workspace-wide questions answered from records the asker can open, with citations, or "not in the records" - Decision log, expertise map (named vs capable backups), 30/60/90 continuation plan, handover rehearsal graded against the evidence, knowledge flagged when its code changes, voice answers in the exit interview, Slack sharing - Eight roles, each with a stated purpose; sample workspaces let you try Relay as any role - Audited handover cases: plan, approve, execute and verify with separation of duties, then follow-through that checks each transfer took hold (measuring the work, not the person) ## Technical - [Architecture](https://github.com/prasadbhalerao1/CodersInPune/blob/main/docs/architecture.md): FastAPI, PostgreSQL with forced row-level security, transactional outbox, Next.js - [API reference](https://github.com/prasadbhalerao1/CodersInPune/blob/main/docs/api-reference.md) and [OpenAPI schema](https://github.com/prasadbhalerao1/CodersInPune/blob/main/docs/openapi.json) - [AI use and guardrails](https://github.com/prasadbhalerao1/CodersInPune/blob/main/docs/ai.md): permission filtering before retrieval, citation validation, injection handling - [Trust: privacy, security, accessibility](https://github.com/prasadbhalerao1/CodersInPune/blob/main/docs/trust.md) - [Testing notes](https://github.com/prasadbhalerao1/CodersInPune/blob/main/docs/round3/testing-notes.md) ## Optional - [Documentation index](https://github.com/prasadbhalerao1/CodersInPune/blob/main/docs/README.md) - [Deployment](https://github.com/prasadbhalerao1/CodersInPune/blob/main/deployment.md)