Skip to content
View mananritwik's full-sized avatar

Block or report mananritwik

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. 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
mananritwik/README.md

Ritwik Manan

Product leader who builds. I run the org and ship the AI myself.

Most recently Head of Product at ScanPay (fintech, 0 to 1, employee #3, through August 2026): I built and led product, design, analytics, and support, and I personally built the AI systems that run them, with Claude Code and Cursor. I'm a product leader who never stopped building. This profile is where I keep the builder side on display: the build-vs-buy calls, how I eval, and where an LLM is dangerous and how I box it in.

Selected work

What The call, and the outcome Role
Support-ops AI stack Ran a support org on a self-built AI stack instead of per-seat vendors, because our SOPs and voice are what a vendor can't sell Owner, 0→1
Call-QA pipeline LLM-as-judge scoring of every call against our real SOPs, at cents per call Built it
Mystic Monk rebuild Rebuilt an agency site from scratch after a malware infection: design, build, and a clean DNS and email migration Built it
UX Proofsheet A Claude skill that benchmarks a design against how real apps ship, evidence-backed and scored Built it
AI-craft notes How I think about evals, agent harnesses, and the AI-era PM bar Author
Sloppify A parody that turns anything into unbearable LinkedIn thought-leadership slop (rules engine + a Claude mode on Cloudflare). I can generate it on purpose because I know exactly what the tells are. For fun

How I think

Reach me

LinkedIn · ritwik.manan@gmail.com

Pinned Loading

  1. support-ai-stack support-ai-stack Public

    Build-vs-buy case study: running a support org on a self-built AI stack (evals, guardrails, system design).

  2. ux-proofsheet ux-proofsheet Public

    Benchmark a UI/UX design against how real apps ship, in your industry and outside it. A Claude Agent Skill.

    Python

  3. mystic-monk-rebuild mystic-monk-rebuild Public

    Case study: rebuilding an agency website after a malware infection (live: mysticmonk.xyz)

  4. sloppify sloppify Public

    Turn anything into an unbearable LinkedIn thought-leadership post. A parody. (rules engine + optional Claude-powered AI mode on Cloudflare)

    HTML