Founding Engineer
In-Person — San Francisco
Gobi is an AI-powered map and messaging product that helps people figure out where to go and what to do with their friends. We imagine a future where Gobi is the app that people use to connect with the real world and the people that matter to them.
Gobi is well funded (in stealth) by top-tier firms and operators. Our team has built legendary, industry-defining products in consumer before.
We’ve built one of the largest structured indices of everywhere (places) and everything happening (events) in the real world. Now, we’re building a clever, social product experience around that index.
Solve problems like this
- Search + ranking at massive scale (200M+ places, millions of events)
- Recommendations + personalization of places + events
- Real-time infra (chat, presence, location)
- AI systems that plan, retrieve, and reason over a multitude of data sources
If you’re excited by systems that sit at the intersection of infra, product, and AI — this is your kind of place.
🔧 The Role
You’ll operate as a founding engineer, owning problems end-to-end.
You will
- Design and ship product + systems from zero → production
- Work directly with founders on product and architecture
- Make decisions that shape the core system (not just contribute to it)
- Move across the stack as needed — frontend, backend, infra, AI
Example problems you might work on
- Ranking what someone should do tonight using embeddings, structured data, and social signals
- Pushing Elasticsearch + vector search to sub-100ms across large datasets
- Building real-time systems that actually feel instantaneous
- Designing map display algorithms that decide what deserves attention
- Building agent pipelines that plan queries and synthesize well-informed answers
- Structuring mountains of invaluable unstructured information on the web to power novel experiences
🧠Areas of Focus
You don’t need to fit neatly into one — most people here won’t.
Application Engineers (Full-stack)
- Build fast, expressive interfaces across mobile + web
- Work deeply on rendering performance and interaction design
- Own features end-to-end (UI → backend)
Typical tools
- TypeScript, React
- React Native / iOS / Android
- Backend systems in TypeScript
Data / AI Engineers
- Build the intelligence layer behind Gobi
- Work on search, ranking, pipelines, and agent systems
- Design systems that operate over large-scale, messy real-world data
Typical tools
- Python and/or TypeScript
- Elasticsearch, Snowflake
- LLMs, embeddings, retrieval + recommendation systems
🧬 What We Look For
We don’t optimize for resumes. We optimize for slope.
Strong signals
- You’ve built something technically ambitious end-to-end
- You have exceptional problem-solving ability (systems, math, or both)
- You move fast and don’t need much structure
- You care about taste — not just if something works, but how it feels
- You like hard, ambiguous problems
Backgrounds we often see
- Competitive programming / math (IOI, IMO, USAMO, ICPC)
- Early engineers at high-bar companies
- Really green github contribution charts
None of this is required — it’s just correlated.
🚫 What This Is Not
- Not a role where you maintain legacy systems
- Not a role with tightly scoped tickets
- Not a role where product is decided for you
If you thrive best with clear boundaries and predictable work, this won’t be a fit.
🚀 Why Gobi
- Real problems — search, maps, AI, and social systems combined
- Small, technical team — everyone ships, including the founders
- High ownership — you can ship meaningful work in days
- Massive surface area — infra, product, and AI all in one system
- Trajectory — the kind of place future founders come from
