AI Researcher (Applied AI & MCP) — Python, Models & Cross-Team Impact
Posted
Description
About Juuli
Juuli is an AI-native fintech platform built exclusively for Europe’s 43M+ solopreneurs. We bring payments, cards, FX, invoicing and tax automation into one place, with 90-second onboarding and AI at the core of the product.
We are a lean team, and every hire here has a direct impact on how we grow.
The Role
You will be Juuli’s AI Researcher. You will test new models as they appear, work hands-on with Juuli MCP, and turn what you learn into practical decisions for our engineering, design and marketing teams. You will work directly with the founders.
What You’ll Do
- Test and evaluate new models as they are released: build repeatable evals on real Juuli tasks and measure quality, cost, latency and safety
- Work hands-on with Juuli MCP: build, test and improve the tools, prompts, agents and workflows that connect models to our platform
- Write Python for prototypes, evaluation harnesses, data analysis and automation
- Turn research into action for Development: agentic workflows, tooling, architecture and ways of building faster
- Turn research into action for Design: AI-assisted prototyping, research and user experience
- Turn research into action for Marketing: content, research, personalisation and workflow automation
- Share what you find through short briefs, demos and playbooks that teams can use right away
- Track the field closely and filter signal from noise: new models, papers, tools and techniques
- Keep experiments safe: data privacy, guardrails and human oversight in a financial context
- Run experiments on our cloud infrastructure and keep costs under control
What We’re Looking For
- 2+ years of experience in applied AI, machine learning or AI engineering
- Strong Python skills: data work, API integration, evaluation code and testing
- Hands-on experience testing and comparing large language models, including building evals
- Experience building with tool use, function calling, agents or MCP
- A solid understanding of what models do well and where they fail: hallucinations, bias, safety and cost
- The ability to turn technical findings into clear recommendations for non-researchers in development, design and marketing
- Fast, curious and comfortable working in a lean team
- Clear writing and excellent English
Nice to Have
- Infrastructure experience on AWS and DigitalOcean: deploying, monitoring and cost management
- Experience with retrieval (RAG), vector search or fine-tuning
- Experience in fintech or another regulated industry
- Public work: writing, talks or open-source projects on AI
- SQL and data analysis skills
What We Offer
- Direct access to the founders and real influence on how Juuli uses AI
- The chance to build our AI research function from the ground up
Skills
Sign in to read and join the discussion.