Are you looking to have an impact on the daily life of millions of entrepreneurs in France (and tomorrow in Europe)? Pennylane is a fast‑growing Fintech dedicated to making accounting and finance simpler for small and medium‑sized businesses.

About Pennylane

We aim to become the most beloved financial Operating System for French SMEs and accounting firms, and soon across Europe. In five years we have built a leading SaaS platform, raised €400 million, grew from 7 founders to 1,000 employees, and served over one million enterprises and 6,000 accounting firms.

Core AI Team – Role Overview

The Core AI team sits at the heart of our AI strategy and builds the shared foundations that power all customer‑facing AI features. As an AI Engineer you will design, implement and maintain the agentic intelligence, quality and models that our solution squads rely on.

Key Responsibilities

- Build and maintain the agent harness – the core runtime that orchestrates LLM calls, tool use, context construction, memory and guardrails – and ensure it is reliable in production.
- Own the evaluation stack: design evaluation pipelines, golden datasets, LLM‑as‑judge and human‑evaluation workflows; track agent performance, regressions and failure modes, and transform production failures into systematic improvements.
- Perform prompt and behaviour engineering and retrieval/RAG at scale, defining what "good" looks like for agents completing complex accounting workflows end to end.
- Host and improve our own small models where appropriate: serve and fine‑tune open‑source LLMs, SFT/DPO/RL training, and document‑extraction models for performance, cost and sovereignty.
- Work closely with Product teams and domain experts (accountants) so that fulfilling real user needs remains at the core of everything we build.

What you can expect

During the first month:

- Learn everything about our company, teams and vision during the first onboarding week.




- Familiarise yourself with our stack and AI tooling (agent harness, evaluation tooling, model providers, MCP). Deliver a few small contributions to get a concrete taste of our tools and processes.
- Meet future stakeholders and gain deep knowledge of our product and operations.

During the first three months:

- Take charge of roadmap items, defining and prioritising tasks autonomously.
- Become comfortable with our technical stack (Python, agentic frameworks, evaluation tooling, model serving and AWS).
- Contribute to larger cross‑team projects.

During the first six months:

- Proactively contribute to the team roadmap.
- Work with engineers and data practitioners on improving our stack, AI platform and evaluation practices.
- Share learnings and best practices within the team.

Beyond six months you will have opportunities to mentor new team members, assume greater accountability in project leadership, and design new processes, tools and best practices.

Qualifications

- 5–8 years of experience; strong in Python and building LLM and agentic systems at scale in production.
- Hands‑on experience with prompting, tool use, context construction, RAG and managing failure, state and reliability.
- Treat evaluation as a first‑class discipline: golden datasets, LLM‑as‑judge, human evaluation, A/B testing, and measuring agent quality, regressions and edge cases.
- Good grasp of applied LLM/ML and AI infrastructure (model serving, vector databases, cost and latency).
- Exposure to model fine‑tuning and post‑training (SFT, DPO, RL).




- Balanced blend of technical, business and product skills; communicate well with non‑technical domain experts and fluent in English (French not mandatory).

We’re looking for someone who

- Speaks English (levels are assessed according to the department).
- Is energized by an ever‑shifting work environment.
- Is highly collaborative internally and with external stakeholders.
- Has enough experience to prioritise business‑led actions in day‑to‑day activity.

Recruitment Process

- First interview with Talent Acquisition Manager.
- Case study interview (75 min).
- Past‑project interview (60 min).
- Interview with Tech & Product leaders (60 min).

Benefits and Perks

- 25 paid vacation days.
- Competitive compensation package.
- Company shares.
- Home office budget and monthly coworking allowance.
- Gym and wellness benefits through Gymlib.
- Language learning via Busuu.
- Latest Apple equipment.
- Remote work allowed within a two‑hour time difference of CET and within Europe.
- Regular company events such as Tech Days and an annual seminar.

For France residents: French contract following French regulation, additional 6–12 RTT, 5 weeks of PTO, lunch credits, health coverage and events in major cities.

Legal and Diversity

Recruitment scam prevention: apply only through official channels, verify sender email domains, and never pay or provide financial information. Pennylane fully embraces diversity, equity and inclusion and provides equal employment opportunities regardless of gender, sexual orientation, origin, disabilities or any other traits.

Personal Data

Pennylane processes your data to manage your application and assess your suitability. If your application is unsuccessful, data may be retained for up to 2 years. You may object at any time and request deletion by writing to

#J-18808-Ljbffr

📌 AI Engineer — Core AI (Madrid)
🏢 Pennylane
📍 Madrid