TL;DR: Build generalised computer-use and a knowledge graph of all the bad people in the world using Python, LLMs, VLMs and traditional ML.

About Us (i.e. Why This Is Going to the Moon)

Sphinx is a well-funded YC startup from the F24 batch. We’re using AI agents to automate the manual work of compliance officers in banks and fintechs – an area where financial institutions spend $30B annually. Our AI agents have already reduced manual reviews by 87% in some cases, all at a fraction of the cost.

Our vision is to build the future of compliance in financial services. We’re using AI agents to turn compliance into just another back-office automation. While doing so, we’re building the largest risk network ever – and setting the stage to solve an even bigger problem: identity. But more on that later.

The Team (i.e. The Dream Team)

Alex (a.k.a. Mr. CEO): Built a system saving Amazon $12M a year as an 18-year-old intern and conducted AI research on LLM hallucinations at MIT. He’ll sell ice cubes to eskimos if he needs to.

Chris (Whiteboard Dude): First employee and Head of Engineering at RelyComply, where he scaled the company from $0 to $12M ARR. Before that, he was a professional DOTA player.

Together, we founded and exited Adorno AI, an audio generation startup, before starting Sphinx.

Role (i.e. What We’re Looking For)

We’re looking for Sphinx’s first founding engineer – this is an incredible opportunity to join a scrappy, driven team taking on an industry that hasn’t evolved since fax machines.

As employee #1, you’ll work directly with us (the founders) from day one to shape the future of the company. You’ll make key technical decisions, build the product, and help assemble a team of A-players.

Here’s What We’re Looking For

• No ego: Company > team > self.
• Love working hard: This isn’t a “chill” job – it will be intense, demanding, and incredibly rewarding.
• Cracked engineer: You write clean, scalable code, solve complex problems, and understand how to build systems that last.
• Client-Driven: We don’t code in the dark. Everything we build is 100% based on client need – we’re here to “build something people want.”
• Ownership: You don’t wait for permission – you see problems, solve them, and own the outcome.
• If this sounds like you, let’s build something great together.

Responsibilities (i.e. What The Hell You’ll Be Doing)

There are two big endeavors Sphinx is working on:

• A Generalizable Robotic Process Automation (RPA) Agent
• Operating in an industry where mistakes cost a minimum of $250k.
• Think Claude Computer Use, but better.
• A Knowledge Graph of All the Bad People in the World
• Leveraging LLMs, graph databases, and website-agnostic scraping.
• Built to handle millions of pages per day at scale.

You’ll be taking the lead on one of these projects, building ML systems from notebook to AWS – essentially the fun part of our current CTO’s work. Tech we’re using to in these projects: Python, LLMs, VLMs, traditional ML, Django, Docker, Postgres, AWS