“Finished products are for decadent minds.”
— Bail Channis, Second Foundation
Hi! 👋 I’m Rohit — software engineer, entrepreneur, and a fan of building things just to see what happens.
How I got here
I grew up in Bangalore, India. My first substantial creation was a website in high school for trading video-game discs — most of my friends would buy a game, finish it, and let it sit on the shelf. The site bridged supply and demand. It hit ~$200 MRR, then turned into an entertainment label with a streamer affiliate program, then a string of in-person gaming tournaments leading to $10,000 ARR. The night I watched ~500 people show up in a room for something we’d organized, that was it. No turning back. Computer science wasn’t a degree choice anymore, it was a tool I needed to be able to keep doing more of that.
I studied CS at the University of Waterloo and used the co-op pipeline to do six internships — deliberately skewed toward pre-seed and seed startups, with one detour into bank fintech and one into research (where I published a small paper). By the time I graduated I knew I wanted to spend my career inside early-stage companies.
In school I joined an accelerator and tried to build a network product that quantified the useful work people had actually done — better than LinkedIn experience. It hit a cold-start problem we couldn’t engineer our way past. The investor conversations that followed taught me the lesson I most rely on: products must solve an immediate problem, not a future one.
What I do now
I’ve spent the last few years inside two very different companies, learning two different things:
- Founding Head of Engineering at Sponsor a Purpose (bootstrapped) — built and scaled the platform from 0 to ~$500k MRR alongside the CEO. This is the school of cost discipline, lean engineering, and shipping under hard constraints.
- AI Engineer at Planned (Series B, scaling toward $10M ARR+) — the school of how mature dev teams structure work, review, and ship at scale. Bootstrapped scaling and Series B scaling turn out to be different skills.
What I’m building in 2025–26
I’m in a “first-principles tinkering” phase, layered on top of the production engineering I learned in the previous chapters. Right now that means:
- Building applications with AI as fast as possible — agentic workflows, prompt libraries, eval harnesses
- Building an LLM from scratch (transformer + training loop + the unglamorous bits) to actually understand what’s under the hood
- Writing Claude Code plugins and personal automations — including a Slack→Linear triage agent that drafts tickets overnight from team signal and only creates them after I approve them by reply, which has changed how I read Slack
- Studying agent orchestration, RAG, multi-agent coordination, and context engineering as a serious craft, not a buzzword
Off the clock
- Game design. I’m building a Slay the Spire 2 character mod themed on Luffy from One Piece. The defining mechanic is a gear-transformation system: each gear gives you escalating buffs in exchange for escalating costs (HP drip, post-exit debuffs, per-combat limits). The fun of it is the design tension — greed vs safety — applied to a card game.
- Reading. Recent stuff: Neuromancer (Gibson opens scenes mid-thought, mid-action — a writing trick I keep stealing), Death of a Salesman, Salt to the Sea. I read across genres on purpose; sci-fi, literary fiction, and history all feed how I think about technology and work. More notes in my library.
How to reach me
GitHub · LinkedIn · X · Discord · email.
For a chronological walk-through of what I’ve built and where I’ve worked, see my timeline.
If you’re working on something in early-stage AI, agent tooling, or a bootstrapped product that needs a founding-engineer brain, I’m always up for a chat.
> "Finished products are for decadent minds."
> — Bail Channis, [*Second Foundation*](https://en.wikipedia.org/wiki/Second_Foundation)
Hi! 👋 I'm Rohit — software engineer, entrepreneur, and a fan of building things just to see what happens.
### How I got here
I grew up in [Bangalore, India](https://goo.gl/maps/TZLVzy6jBxneBGMZ7). My first substantial creation was a website in high school for trading video-game discs — most of my friends would buy a game, finish it, and let it sit on the shelf. The site bridged supply and demand. It hit ~$200 MRR, then turned into an entertainment label with a streamer affiliate program, then a string of in-person gaming tournaments leading to $10,000 ARR. The night I watched ~500 people show up in a room for something we'd organized, that was it. No turning back. Computer science wasn't a degree choice anymore, it was a tool I needed to be able to keep doing more of *that*.
I studied [CS at the University of Waterloo](https://uwaterloo.ca/) and used the co-op pipeline to do **six internships** — deliberately skewed toward pre-seed and seed startups, with one detour into bank fintech and one into research (where I published a small paper). By the time I graduated I knew I wanted to spend my career inside early-stage companies.
In school I joined an accelerator and tried to build a network product that quantified the *useful work* people had actually done — better than LinkedIn experience. It hit a cold-start problem we couldn't engineer our way past. The investor conversations that followed taught me the lesson I most rely on: **products must solve an immediate problem, not a future one.**
### What I do now
I've spent the last few years inside two very different companies, learning two different things:
- **Founding Head of Engineering at [Sponsor a Purpose](https://sponsorapet.org/)** *(bootstrapped)* — built and scaled the platform from 0 to ~$500k MRR alongside the CEO. This is the school of cost discipline, lean engineering, and shipping under hard constraints.
- **AI Engineer at [Planned](https://planned.com/)** *(Series B, scaling toward $10M ARR+)* — the school of how mature dev teams structure work, review, and ship at scale. Bootstrapped scaling and Series B scaling turn out to be different skills.
### What I'm building in 2025–26
I'm in a "first-principles tinkering" phase, layered on top of the production engineering I learned in the previous chapters. Right now that means:
- Building applications with AI as fast as possible — agentic workflows, prompt libraries, eval harnesses
- Building an LLM **from scratch** (transformer + training loop + the unglamorous bits) to actually understand what's under the hood
- Writing **Claude Code plugins** and personal automations — including a Slack→Linear triage agent that drafts tickets overnight from team signal and only creates them after I approve them by reply, which has changed how I read Slack
- Studying agent orchestration, RAG, multi-agent coordination, and context engineering as a serious craft, not a buzzword
### Off the clock
- **Game design.** I'm building a [Slay the Spire 2](https://www.megacrit.com/) character mod themed on Luffy from One Piece. The defining mechanic is a gear-transformation system: each gear gives you escalating buffs in exchange for escalating costs (HP drip, post-exit debuffs, per-combat limits). The fun of it is the design tension — greed vs safety — applied to a card game.
- **Reading.** Recent stuff: *Neuromancer* (Gibson opens scenes mid-thought, mid-action — a writing trick I keep stealing), *Death of a Salesman*, *Salt to the Sea*. I read across genres on purpose; sci-fi, literary fiction, and history all feed how I think about technology and work. More notes in my *[library](/library)*.
### How to reach me
[GitHub](https://github.com/rkaushik29) · [LinkedIn](https://www.linkedin.com/in/rkaushik29/) · [X](https://x.com/rohitkaushik29) · [Discord](https://discordapp.com/users/695689515506860082) · [email](mailto:rohit.rk.rk1@gmail.com).
For a chronological walk-through of what I've built and where I've worked, see my *[timeline](/work)*.
---
*If you're working on something in early-stage AI, agent tooling, or a bootstrapped product that needs a founding-engineer brain, I'm always up for a chat.*