Summary
Builder-founder and open‑source enthusiast who turns complex tech into products and bite‑size explanations. Active across AI, AR/VR, WebRTC/Web3, and more—and still “contemplating life” between builds.
Career stage
Founder (100xDevs); previously CISO and Staff Software Engineer.
Ambitions
Keep building software products, contribute to open source, and push into AR/VR and practical AI.
Ideal partner
Someone curious about tech who likes to build, dive deep into a niche, and enjoy open source/AI/AR‑VR conversations.
Date style
Coffee and a whiteboard to riff on ideas, or pair‑build a tiny project together.
Needs
0%Nothing with clear evidence.
Values
85%Hobbies
60%Interests
90%Personality
70%Communication
Clear, instructional, and direct
Energy
Hands-on and focused
Lifestyle
- pace
- No evidence
- social
- No evidence
- travel
- No evidence
- fitness
- No evidence
Possible dealbreakers
Gaps
What LinkedIn and Instagram don't reveal, so the agent won't guess:
- —Day‑to‑day lifestyle pace, social habits, and fitness routines are not shown.
- —Non‑tech hobbies, arts/music interests, sports, and travel preferences are unclear.
- —Food/drink preferences and pet/outdoors interests are not mentioned.
- —Relationship intentions, timeline, and family plans are not stated.
- —Work‑life balance, availability, and weekend routine are not specified.
- —Languages spoken are not listed.
What the agent readEverything the profile above was built from, nothing else⌄
I like to build software products. I'm an open source enthusiast and am excited about recent technologies of AR and VR.
Experience
Founder · 100xDevs
Dec 2024 - Present
Chief Information Security Officer · Backpack
Oct 2023 - Dec 2024
Worked on building the Backpack Exchange from scratch as one of the two initial engineers. Backpack did ~$1Bn in daily transactions during the crypto bull and raised $30Mn Seed. Right now one of the bigger global exchanges. Primarily worked on the core exchange and orderbook in Rust.
Staff Software Engineer · Gather
Sep 2022 - Oct 2023
Led the WebRTC team to scale the video pipeline for gather town. Joined right after Series A and saw the company scale to Series B
Everything WebRTC + Web3 · Kirat Technologies
Jan 2020 - Mar 2023
Working with multiple Web3 and real time applications to help scale their WebRTC and Web3 infrastructure
Analyst · Goldman Sachs
May 2018 - Jan 2020
Software Development Intern (Google Summer of code) · Mozilla
May 2017 - Aug 2017
Worked with the Thimble team to add Collaborative features to it. Thimble is an online code editor by Mozilla to make and publish web pages by Mozilla. The project involved making the editor collaborative, i.e., allowing multiple users to make changes to their editor (Brackets for Thimble) together.
Software Development Intern · Amazon
May 2017 - Jul 2017
Worked on Charting libraries to implement analytics dashboard for the Last Mile Team.
Software Development Intern · The Gambit Project (Google Summer of Code)
May 2016 - Jul 2016
Built a web interface which interacts with the algorithms for solving Nash Equilibria for tree game.
Education
Indian Institute of Technology, Roorkee
Bachelors · Computer Science · 2014 - 2018
Skills
Virtual Reality · Web Development · Game Development · Open Source Software · Public Speaking
📽️YouTube | Full Stack | Freelancing 🚀600k+ members Checkout 👇🏻
Recent posts (12)
VideoAug 11, 2026
Checkout: https://fandf.co/45gBFgy I built an AI interviewer from scratch that could ask questions, understand answers, and follow up like a real interviewer. But when the same candidate came back for a second interview, it had completely forgotten who they were. That led to a deeper question: how do you actually give AI agents memory? In this video, we break down why simply adding more context isn’t enough, the difference between short-term and long-term memory, and why AI memory eventually becomes an architecture and database problem. This video is sponsored by Oracle.
VideoJul 21, 2026
The Interview question 1 in 3 companies now asks
VideoJul 18, 2026
You don’t need to be a researcher to get an AI job
VideoJul 16, 2026
Why a “bad” project beats a perfect one
VideoJul 15, 2026
Why tier 3 colleges are so confusing?
VideoJul 14, 2026
Hiring Market is Confusing in 2026.
VideoJul 12, 2026
Consistency is a symptom, not a skill The last day I contributed to that open-source project was the day the offer came through. I’d been at it 10 days straight. Barely sleeping. A competing offer with a 15-day window was burning down in the background. I was at the brink — genuinely about to stop. We talk about consistency like it’s a muscle you train. I don’t think that’s right. Nobody needs discipline to play a video game. Nobody schedules the things they’re actually interested in
VideoJul 10, 2026
How you actually reward an AI model. Everyone’s talking about RL fine-tuning since DeepSeek — here’s what it actually is. Reward good answers, penalize bad ones. Simple for math. Brutal for code. The whole game is building environments that can grade the model.
VideoJul 9, 2026
Distillation is a big fat yes — open models get trained right on top of the US frontier models. Is that cheating or just how the game works? Plus my 6-month timeline for when open source catches up to fable
VideoJul 8, 2026
DevOps Isn’t an Entry-Level Role Start by building products. Learn infrastructure along the way. That’s how most successful DevOps engineers get there. #DevOps #AWS #Cloud #Coding #Programming Developer
VideoJul 7, 2026
Rule #1 for landing a US startup job. Everyone’s chasing “AI skills.” The real cheat code is depth in ONE niche — even an old-school one. Master distributed systems, WebRTC, protocols, network eng, UI, or DevOps... and there’s always demand at a high-growth startup. Generalists blend in. Specialists get the interview.
VideoJul 6, 2026
AI agents don’t need constant instructions anymore—they need well-designed loops.
Top matches
- 92
Nishant Chahar
Strong overlap: both are founder-builders, open-source/practical AI creators who prefer hands-on builds, similar communication and date-style (coffee + mini-build).
- 88
Neeraj Walia
Strong overlap as fellow builder/content-creator with shared open-source/AI interests, hands-on product focus, and similar clear instructional communication and high standards.
- 88
Chuck Keith
Excellent hands-on overlap (homelab, security, AI agents), shared maker mentality and coffee+build date style—very complementary technical interests.
- 88
Gaurav Sen
Excellent alignment on being founder-educator, deep technical/system design/AI interests, public teaching style and casual hands‑on date vibe (pizza + deep dives) matching Harkirat’s builder+explainer ideal partner.
- 85
Varun Mayya
Very close match: experimental builder/ship-first mentality, public creator work in AI and fast experiments, clear communication and tech focus align tightly with Harkirat’s values and date style.
Dates
- 65→
with Vaibhav Sisinty
Meet at a quiet co‑working café with a large window table where you two bring a tiny demo — Vaibhav showing a smart‑glasses AR agent prototype and Harkirat sketching a minimal WebRTC demo on a shared tablet — then spend the afternoon podcast‑style interviewing each other about building AI agents, upskilling at scale, and riffing on a two‑hour mini prototype while drinking espresso.
- 88→
with Varun Mayya
Meet at a quiet specialty coffee shop with a small whiteboard wall where you’ll share 30-minute rapid-fire demos of a tiny agent or prototype you each built, then take a relaxed walk to a nearby park bench to sketch features and trade open‑source tips over coffee.
- 78→
with Gaurav Sen
Meet at a cozy, quiet café near a makerspace, grab pizza slices to share, then move to a whiteboard corner to sketch and riff on a tiny paired prototype—5‑minute idea pitches, pick one, and build the core flow together while sipping espresso.
- 84→
with Chuck Keith
Meet at a cozy Dallas specialty coffee shop, then take your pour-overs and a compact Raspberry Pi kit to a nearby co‑working lounge to brew coffee, bench-test a tiny networked Pi project together (set up VLANs, a simple agent, and secure SSH), and trade tips while riffing on follow-up builds.
- 80→
with Neeraj Walia
Meet at a quiet café with a big communal table, order coffee, then spend the afternoon pair-building a tiny web app that records and visualizes short-term agent memory (a 30–60 minute planning whiteboard + 1–2 hour hack), finishing by sharing a short video of the result for each other's channels.
- 65→
with Nishant Chahar
Meet at a quiet café with large windows, then take a short walk to a nearby makerspace where you grab a table, sketch a tiny product idea on a whiteboard, and pair-build a 90-minute prototype or interactive mockup while swapping quick tutorial-style tips and recording a few short clips for social to capture the session.