Skip to content
View samarpassey's full-sized avatar

Block or report samarpassey

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
samarpassey/README.md

Samarpartap Passey

AI developer building agentic systems — MCP servers, LLM evals, and voice agents that run in production.

maple-procure — An MCP server over Canada's public tender and award data. Five tools across stdio and streamable HTTP, scoped bearer auth, a structured log line per call, and a write that takes two calls. Deployed, with a read-only demo token in the README so you can call it yourself.

ledger — Claim-level grounding eval for RAG. Verifies each claim against only the chunk it cites, and finds the failure mode citation metrics can't see.

all-quote — Ontario auto-insurance quote agent. One intake, 79 rate sources, an evidence-backed terminal status for every one.

Sage — Voice-first AI interview coach running in production. Claude Sonnet 4.6 plans and scores; Vapi carries the live voice.

Python · TypeScript · MCP · FastAPI · Next.js · Anthropic SDK · Playwright

📍 Toronto, Ontario · samarcodes.com

Pinned Loading

  1. mapleprocure-triage mapleprocure-triage Public

    Hourly n8n workflow that triages Canadian federal tender notices with Claude and routes everything uncertain to a Postgres review queue.

    JavaScript

  2. maple-procure maple-procure Public

    An MCP server over Canada's public tender and award data (CanadaBuys). Scoped auth, provenance on every result, and a write that takes two calls.

    Python 1

  3. ledger ledger Public

    Claim-level grounding eval for RAG. Verifies each claim against only the chunk it cites, and surfaces the failure mode citation metrics can't see.

    Python 1

  4. all-quote all-quote Public

    Personal-use Ontario auto-insurance quote agent. One intake, 79 distinct rate sources, evidence-backed terminal status for every one.

    Python

  5. Sage Sage Public

    Voice-first AI interview coach in production. Live spoken interviews, Claude for planning and scoring, short-lived scoped JWTs, Clerk, Supabase.

    TypeScript

  6. careerbytes-workshop careerbytes-workshop Public

    Workshop I taught on building your own AI interview agent — full build guide, no coding experience required.

    1