San Francisco · Seed Stage · Hiring Now

Careers at Kiloforge

Building world-class software for the millions of specialized communities that are underserved.

Backed by

a16z Facebook, GitHub, Coinbase

Uncork Capital Fitbit, Postmates, PagerDuty

YC Partners Airbnb, Stripe, Dropbox

Rahul Vohra Founder, Superhuman

Founding Team

Who We Are

Google Brain DeepMind Facebook Microsoft Harvard CS YC S20 a16z

We're AI researchers from Google Brain and DeepMind, engineers from Facebook and Microsoft, and graduates of Harvard CS and Applied Math.

We founded Start Playing Games (YC S20, a16z-backed) and scaled it to profitability with $60M+ transacted. We built and sold Clientelligent, an AI company acquired by Canoe Financial in 2022.

Now we're building Kiloforge.

Vision

What is Kiloforge?

We're building an autonomous app factory — a system that discovers opportunities and user pain points across millions of communities, validates ideas, designs products, writes code, launches to market, acquires customers, handles support, and evolves through adaptive learning.

Our vision: 10,000 self-evolving micro-apps, each solving real pain points for specialized communities, each generating $1M in annual revenue.

We're building for the future — a future where AI directs entire companies. We're making that future real, today.

Culture

"We hold two truths — we are excellent and we need to get better."

Excellence is asymptotic — we may never arrive, but that should never stop us from climbing higher.

Principles

Two Rules

Everything at Kiloforge flows from two foundational principles. They are the lens through which we evaluate every decision, every hire, and every line of code.

Proactivity

The transition from good to great follows the same pattern: doing the same thing, but proactively. It's the difference between accepting feedback when it comes versus actively seeking it out. Between fixing problems when they become critical versus preventing them before they happen.

Proactivity is a bias toward action and anticipation. It means not waiting to be asked, not waiting for permission, and not waiting for the perfect moment. It means seeing what needs to be done and doing it — before anyone else notices it needs doing.

Good Judgment

Proactivity without wisdom is recklessness. Good judgment means applying calibration and careful thinking to every action. It means knowing when to push and when to pause, when to act alone and when to collaborate.

Good judgment is the counterbalance to speed — the compass that keeps proactivity pointed in the right direction. It's the difference between moving fast and moving fast in the right direction. These two principles together form the foundation for how every value is lived at Kiloforge.

Product

See What We're Building

Open Positions

Join the Team

Ship one micro-app per week, each better than the last. Build the engineering pipeline for deploying real apps to production and help achieve $1M ARR on at least one portfolio app by year end.

Compensation $180K – $225K

Equity 1.00% – 1.50%

Experience 5+ years building & scaling production software

Stack TypeScript, React, Node, Python, SwiftUI

Non-negotiable

Must have built and shipped production software used by thousands of real users — not side projects, but products you owned and scaled professionally.

Apply Now

Culture

Life at Kiloforge

Latest Hardware & AI Tools

Top-of-the-line equipment and subscriptions to every AI tool you need.

Health, Dental & Vision

Comprehensive coverage so you can focus on building.

Flexible PTO

Take the time you need. We trust you to manage your own schedule.

San Francisco

In-person collaboration in the heart of the tech ecosystem.

FAQ

Common Questions

How do you identify which apps to build?

We run a systematic niche discovery process — combining external trend data, community signal mining, a network of taste makers embedded in niche communities, and rapid 48-hour validation sprints. We score every opportunity on pain intensity, willingness to pay, community density, and cross-promotion fit. If a niche scores high, we build. If it doesn't, we move on fast.

How does quality compound over time?

Every app we ship improves the quality baseline for the next. Reusable components, proven UI patterns, tested deployment scripts, and fine-tuned AI prompts all compound. App #1 starts at 72% quality. App #10 starts at 85% and ships 80% faster. Every hour spent improving our scaffolding saves ten hours across future apps.

Why niche markets instead of going broad?

Unit economics. Niche CAC runs $5–15 through community-driven, word-of-mouth growth versus $50–150 for broad paid acquisition. Passionate users churn less and pay more. A $50M niche is a rounding error for Google but a generational outcome for a lean operation. Sales and marketing spend can be up to 50% lower in vertical software versus horizontal.

How is Kiloforge different from Lovable, Bubble, or other AI builders?

Those are tools for other people to build apps. We build and own the apps ourselves. We're not a platform — we're operators. We capture 100% of the value we create instead of taking a platform fee. Our cross-promotion network, compounding data flywheels, and deep niche expertise create moats that tool platforms can't replicate.

What happens when AI can build apps instantly?

Even if AI builds apps instantly, someone needs to decide what to build and for whom — that's our core competency. Our apps will already have users, data, and community moats. We either win because our portfolio is entrenched, pivot to become the curation layer for AI-generated apps, or become an acquisition target for AI platforms that need our niche expertise.

What if some apps fail?

Expected and planned for. We assume 50%+ of apps won't reach scale. Our cost structure means failures run $5–20K each — not hundreds of thousands. Every "failure" improves our scaffolding and knowledge. We only need 2–3 winners out of 10 to hit targets. The portfolio approach is the strategy, not a fallback.

What does the tech stack look like?

We start with off-the-shelf foundation models (Claude, GPT-4) and focus on product-market fit. As we identify bottlenecks, we build proprietary tooling around them — custom RAG systems, niche-specific UI components, automated QA, and a cross-promotion engine. The moat isn't the model. It's what you build around it.

What's Kiloforge's unfair advantage?

Three things. Founder-market fit — we built and scaled Start Playing Games in an adjacent space, so we know niche communities. Technical leverage — AI-native from day one, not retrofitting. And portfolio strategy — every app strengthens the next through cross-promotion, shared infrastructure, and compounding data. Most builders go one-at-a-time. We build the factory.