Krithik Sai Sreenish Gopinath
I'm a product manager who ships — I bridge customers and engineering, find what's actually blocking progress, and either build the fix myself or get the right people moving on it.
Product Associate at Matic Robots. Own the crown-module roadmap (cameras, infrared, microphones). Cut false-positive rejection rates from ~50% to ~32% in three months.
- PM at Saayam (consumer mobile). Shipped the in-app chat coordination feature that cut task-completion time ~20% post-launch.
- PM intern at Wurq. Owned the ML pipeline for a pose-correction feature. Inference latency 490ms → 67ms; edge costs down ~28%.
- Ops at ZS Associates. Owned post-launch optimization for a $32M pharma platform. Killed ~35% of support volume. Protected ~$500K in billable time, enabled ~$400K in new revenue.
Writing a search engine in Rust from scratch. Last week I wrote my own encoder to shrink the search index by 76% — and it also got 50% slower (I can tell you exactly why). Built filmsearch, a semantic movie search over 54K titles using Llama 70B, where my ranking beat the baseline by ~35%. Built voyantra, a trip planner on the Amadeus API.
A startup where I can sit between customers and engineering and translate each side to the other. Specifically: large fragmented industry, low NPS, room to vertically integrate — the Keith Rabois formula.
M.S. Engineering Management, Northeastern University · B.Tech. CS, VNIT Nagpur.
Hi, I'm Krithik.
I love search engines, non-fiction, the Hindu epics, Carnatic music, and the Denis Villeneuve filmography. I build things in my free time, because I found out how fun it is. Right now I'm at Matic Robots, building robots.
The free-time projects are where my head actually lives. I'm halfway through writing a search engine in Rust, working through Manning's Information Retrieval textbook one chapter at a time. Last week I replaced a serialization library with my own variable-byte encoder and shrank my index by 76%. It also got 50% slower — and I can tell you exactly why.
Before the Rust engine I built filmsearch, a search engine for movies that understands queries like "films about love and space." A founder used it to find a Japanese animated film he couldn't find anywhere else. Still my favorite piece of feedback.
Before that, voyantra — a trip planner where I figured out what an API was, what a backend does, what a database is for. Nothing in it is novel. But it's the project that taught me how to build the things that came after.
And before any of that, a marketing agency in undergrad. $4K a month, seven clients. I shut it down to graduate. It taught me more about how businesses actually work than any class I took.
At Matic I own the roadmap for the crown module — cameras, infrared, microphones. In three months we cut false-positive rejection rates from ~50% to ~32% by diagnosing root causes on the production line and tightening the acceptance bounds with engineering.
Before Matic I was a PM at Saayam (consumer mobile) and a PM intern at Wurq (ML, on-device inference). Before that, ops at ZS Associates. Details below. I trained as an engineer at VNIT in India, then did a Master's in Engineering Management at Northeastern.
Looking for the next thing. A startup where I can sit between customers and engineering and translate each side to the other. Specifically: large fragmented industry, low NPS, room to vertically integrate — the Keith Rabois formula.
what I'm doing now
- owning the crown-module roadmap at Matic — cameras, infrared, microphones. Coordinating across hardware, firmware, and ML.
- writing a search engine in Rust, chapter by chapter through Manning's Information Retrieval textbook. Currently on Ch. 6 — TF-IDF and ranking.
things I've built
A search engine for films that understands queries like "movies about love and space." Custom BM25 from scratch, query relaxation, sentence-transformer zone selection. End-to-end: Scrapy → Postgres → React on Vercel. A startup founder used it to find a Japanese animated film he couldn't find anywhere else — that's still my favorite piece of feedback. code
Working through the Manning IR textbook one chapter at a time, building everything from scratch in Rust. Currently: a positional inverted index over 12K documents with my own VByte encoder. Next: TF-IDF ranking. The goal is to actually understand what Lucene does, not just use it.
Started in undergrad. Got it to $4K/mo with 7 clients. Ran the whole show — Facebook and Instagram ads, sales funnels on ClickFunnels, sales calls, delivery, billing. Closed it to graduate. The most useful thing I've ever done — every later job I've had makes more sense because I had to do payroll once.
A trip planner with real flight and hotel data via the Amadeus API. MERN stack, 100+ commits. This is the project where I figured out what an API actually was — what a backend does, what a database is for, how the frontend talks to either of them. Nothing here is novel; everything here taught me how to build the things that came after. code
A one-day AI hackathon project at the SundAI Club at MIT. The team built a tool that takes hours of conversation audio, finds the moments people laughed, and uses an LLM to summarize the joke or punchline that landed. I was one of about sixteen hackers — turned up, contributed where I could, shipped it inside a day. The kind of thing that reminds you how fast a small team can build something real.
Own the roadmap for the crown module on the production line — cameras, infrared, microphones. Coordinated across hardware, firmware, and ML teams to define acceptance bounds and tune thresholds against human-interaction baselines. Cut false-positive rejection rates from ~50% to ~32% over three months by diagnosing root causes in production QA, unblocking throughput on the line.
PM intern. Owned the ML pipeline for pose correction on a fitness app — IMU sensor data, training data collection, requirements for a CNN. Inference latency went from 490ms to 67ms; on-device edge processing cut per-inference costs by ~28% against the cloud baseline. The investor liked the work enough that I kept building on his project after the internship ended.
First job out of undergrad. Owned post-launch optimization for a $32M enterprise platform that pharma sales reps used to distribute product. Triaged support tickets, found the patterns, shipped the fixes — eliminated ~35% of support volume over two quarters. Made the case to leadership for optimization over new features, protecting ~$500K in annual billable time and enabling ~$400K in new feature revenue.
the through-line
Every project above did the same thing: helped someone solve something that actually mattered to them. The pharma rep needed the platform to just work so she could go sell. The film fan needed the search to understand what she meant, not what she typed. The robot needs to tell the kid from the toaster.
What I bring is people, engineering, and clear thinking — pointed at problems that matter. I figure out what's actually getting in the way, get the right people moving in the same direction, and either build the thing myself or work alongside the people who do.
other things
I read a lot, mostly non-fiction. Right now: Ramesh Menon's modern rendering of the Mahabharata. Recently finished his Ramayana and Krishna: The Blue God. Last year I went through Walter Isaacson's The Innovators and Neil Postman's Amusing Ourselves to Death. Full list on Goodreads.
Played chess for the team at VNIT and at Northeastern. Still play, badly, online.
stuff I think is pretty cool
Italian painter Giampaolo Tomassetti's Mahabharata series. I keep coming back to Parthasarathi — Krishna as Arjuna's charioteer on the eve of Kurukshetra, the exact moment before the Gita. An Italian painter rendering this scene with Caravaggio's light is the kind of cross-cultural crossover that shouldn't work and does.
A recent paper on raga music and emotional regulation. The premise — that specific ragas in Indian classical music act on specific emotional states in reproducible ways — is something I'd heard about from family for years and always filed under folk wisdom. Turns out there's a literature. I'm still working through it.
the contrarian question
Peter Thiel has a famous interview question: what important truth do very few people agree with you on? It's hard because most people either can't answer it (no original thinking) or dodge it (too afraid to commit in public). I want to see what you've got.
Anonymous. I read everything. The best ones show up below, on this page, without your name attached.