“This is the most exciting time in history to build.” Sam Altman
Should I start a company? Should I keep going with my startup? What problem should I solve?
“Starting by thinking up a problem is risky because you may invent a problem that doesn’t truly exist. If you can’t answer [‘who wants this right now?’], then the idea is probably bad.” Paul Graham
We believe the best thing you can do right now is join a breakout AI team, face real problems, work with potential co-founders, and make a lot of money.
We’re solving the biggest problem in modern work: email.
And we’re doing it at the best possible time in history to solve it.
Too many companies start out focused on what users love, then add bells and whistles no one needs, hide features behind paywalls, and lose sight of what made them great in the first place.
We think the best user experiences come from focus.
For us, that means AI-generated email drafts. We’re laser-focused on making them better over time, using context from your meetings, inbox, and messages, because that’s what matters most to our users.
Unlike other startups, we’re a pure-play applied AI company, not a SaaS tool with AI bolted on. We use the best techniques, including fine-tuned open-source models, tool use, and retrieval-augmented generation. As a result, users send 53% of the drafts we generate.
Your Background:
5+ years of experience building and shipping production software, with deep expertise in Python or TypeScript and a track record of owning systems end-to-end.
Hands-on experience with modern AI/ML tooling (LLMs, fine-tuning, RAG, embeddings, distributed training, inference optimisation, etc.) and proven ability to work directly with models in production.
Strong product intuition and user-centric mindset, able to translate ambiguous problems into simple, high-impact product decisions without product managers.
Startup-ready mindset: comfortable operating with extreme ownership, rapid iteration, and high autonomy within a small, elite technical team.
How’s it going?
This year, we’ve grown from $1M to $25M ARR, with a technical team of just 12 (up from 2).
We’re forecasting $100M ARR by the end of 2025.
We have product–market fit, and we take none of it for granted.
We’re relentlessly focused.
There will be a $10B+ company in this space.
It will be us if we stay focused.
Ranked 7th in Andreessen Horowitz’s list of top AI apps, based on real startup spend data. The other nine in the top 10 are valued at $10B+.
Team members own both product strategy (we have no product managers and no plans to add them) and the execution of that strategy.
They decide when to bring in qualitative insights (interviews, surveys, etc.) or quantitative data, supported by our data engineering team.
Our technical staff focus on:
Building an incredible product
Helping users find value and share
Training cutting-edge models
Developing scalable infrastructure
A small team of exceptional people, working hard on a problem they care about, will always beat a larger, less focused team.That intensity means this isn’t a normal role, but it also means you’ll move faster than anywhere else.
Examples:
Serafin joined 6 months ago and built our entire product notetaker, competing with whole companies, including research, model training, and product development.
Joey joined 7 months ago and built our human data platform for model training and evals.
Josh joined 14 months ago and now leads infrastructure, having built core products like scheduling along the way.
Most fast-growing startups fall into traps: overhiring, losing focus, acting like big companies.
We believe small, exceptional, high-performing teams are the future.
No bureaucracy
No processes that slow decisions
No constant distractions, we have company-wide focus time with no meetings until 3pm
A humble but extremely ambitious environment.
A group of people giving unreasonable effort to the right things.
What we payWe offer the highest cash and equity in London, up to £500k package!
#LI-FYXER
Top Skills
Fyxer AI London, England Office
Chancery Lane, London, United Kingdom
.png)
