AI-Native / AI-Forward Developer
Software Engineering, Data Science
Remote
About IMPOSSIBLE
IMPOSSIBLE is a crypto investment platform. We focus on high-quality, fair, accessible opportunities for people around the world.
We try to make DeFi easier to use: investing more transparent, trading cheaper, and yield easier to reach. Through our accelerator and launchpad, we help new crypto projects come to market and give users deals we have actually looked at.
You will work across the group: IMPOSSIBLE, Rarible, and our AI company Impossibuild. The work sits in AI, crypto, DeFi, marketplaces, and whatever we build next.
The Role
We want an AI-native developer. Someone who uses AI as a serious way to ship, not someone who pastes a question into ChatGPT once in a while.
You have changed how you build around coding agents and modern tools, and you still have your own judgment. You can tell when the model is helping and when it is making a mess.
AI can write the code. You still own the engineering.
- Break a hard problem down and give the model the right context.
- Use coding agents on larger pieces of work, not just one-off snippets.
- Steer them when they assume the wrong thing or wander off.
- Read and challenge the code they produce.
- Debug it, test it, and check that it actually works.
- Make the architecture and product calls yourself.
- Know when AI is useful, and when you should just write it.
- Move faster without getting sloppy on security, reliability, or the next person who has to maintain it.
What you'll work on
Depending on the project, you might sit in IMPOSSIBLE, Rarible, Impossibuild, or another company in the group.
- Ship AI-powered product features.
- Build internal agents and tools that save people time.
- Put LLMs into frontend and backend systems.
- Build AI workflows for research, risk, trading, and ops.
- Work with wallets, smart contracts, and on-chain data.
- Build around DeFi, marketplaces, trading, and crypto infrastructure.
- Use AI to personalize how people use the product.
- Build retrieval systems, data pipelines, APIs, and automations.
- Try new models, agents, frameworks, and APIs.
- Improve prompts, context, evals, latency, cost, and reliability.
- Prototype quickly and ship new AI-native products.
What we're looking for
Shipping real products matters more than matching every item on this list. Crypto, blockchain, or DeFi experience is a plus. Strong engineers who can pick up the domain are welcome too.
Engineering
- Python
- TypeScript / JavaScript
- Go
- React / Next.js
- Node.js and backend frameworks
- APIs and services
- SQL and modern databases
- Cloud infrastructure
- Git, CI/CD, testing, monitoring, and building things that are safe to run
AI tools
- LLMs and LLM APIs
- AI coding agents
- Claude Code, Codex, Cursor, Copilot, or similar
- Agent frameworks and tool calling
- Vector databases and RAG
- Prompt and context engineering
- Eval and observability
- MCP and agent workflows
Proof of work, not just a resume
We care more about what you have built than where you worked or what degree you have.
Send projects, GitHub repos, live products, demos, open source, experiments, or write-ups. Then answer these three questions in the form.
- 1.Show us something you built where AI clearly changed how fast or how well you could do it. What tools and workflow did you use? What did the model do well, and what did you have to steer, fix, or rethink?
- 2.Tell us about a time an AI coding tool gave you an answer that looked right and was wrong. How did you catch it? What was actually wrong? What did you change after that?
- 3.If you had 48 hours to ship a production feature, how would you use AI from idea through architecture, implementation, review, testing, deploy, and monitoring?
How to apply
Click Apply now. Include your resume, proof of work, and answers to the three questions above. Extra credit if the application itself shows how you think.
We read every application. We only write back to people we want to take further.