Full-stack Engineer (AI)

Better Trips for travelers. Better Trips for the World

Since Evaneos was founded in 2009, we’ve been building a greener, fairer future for travel. By connecting travelers with the world’s best local experts, we provide trips that are better for them and better for the places they visit. For us, it’s all about quality and impact. That means fighting mass tourism and impersonal experiences at the expense of local communities and the environment. Instead, we offer tailor‑made trips, lovingly crafted by those passionate about sharing and protecting the destinations they call home.

Thanks to our network of more than 600 local agencies in more than 160 destinations, we’ve helped more than 800,000 travelers plan extraordinary trips worldwide. The adventure doesn’t stop here. We invest a percentage of every trip we sell in our Better Trips Fund and have committed to spend at least 3 million € by 2025 on projects that reduce the negative impacts of travel and drive positive change within our industry. Think: creating new opportunities for woman‑owned businesses, guiding small businesses through sustainability training and certification, and much more.

Our values are just as important within Evaneos. Our strength is in our diversity, made up of a multitude of languages, backgrounds, and experiences. We believe in building the best teams by creating an environment where everyone can be authentic, trust each other, and thrive.

We hope you’ll join us on our mission!

You can also follow our Glassdoor page and see the comments of our current employees and candidates.

At Evaneos, our mission is simple: to help travelers experience trips that are more local, more responsible, and more human, all around the world.

The innovation Department

Within the Innovation Department, the Pathfinder team is reinventing how travelers design their tailor‑made trips, with AI as the central lever of the experience.

We are hiring an AI Product Engineer to ship AI‑native travel experiences used by real travelers in production. You’ll design, build, and operate AI products powered by agentic architectures, working end‑to‑end from problem framing to live monitoring, always in service of real user and business needs.

Your impact: you’ll build traveler‑facing AI products that move metrics on conversion, experience, and cost. Two are already live and serving traffic: an AI trip designer that helps travelers explore and customize itineraries and a conversational experience that turns their intent into a structured travel project local agencies can act on. Real products, not internal tools.

Our stack, at a glance:

• Languages: Python, TypeScript
• AI/LLM: LiteLLM Gateway, Langfuse, Langsmith, multi‑provider (Anthropic, OpenAI, Gemini, Mistral), Langchain DeepAgents, Claude Agent SDK
• Design agent architectures: runtime, tool systems, orchestration patterns, middleware, state management, RAG.
• Own context engineering: shape what the models see and find ways to better steer the model and system, raising quality while controlling cost.
• Make quality measurable: build evaluation pipelines (offline benchmarks plus online metrics) and observability to catch hallucinations, drift, and regressions before users do.
• Own the economic viability: once something is live, find the right balance of quality, speed, and cost.
• Keep it running: debug production issues on non‑deterministic systems, handle model migrations when providers change their APIs, and deal with structured output failures and tool call errors.
• Own the infra: you build it, you ship it. Keep the dev environment fast and the deployment pipeline reliable.
• Own problems end‑to‑end: from framing to resolution, in collaboration with product, design, and business stakeholders.

Be a bridge between tech and product. Take an active part in product direction, grounded in real user and business stakes. Translate technical constraints into product options, and explain the paradigm shifts AI brings. Proactively communicate on progress, decisions, and blockers: position yourself as an enabler for the team.

Grow our internal AI community. Co‑lead the AI Champions Tech program, leveling up Product & Tech across Evaneos on AI practices and knowledge. Share your learnings with the Innovation team and wider tech community. Run workshops and talks that demystify what matters in the agentic era.

You build the product, not the platform. Focus on shipping AI experiences directly to travelers: real users, real conversion, real business impact.

Full ownership in a team that fits in one room. No ticket queues, no handoffs between ML and product teams. You frame problems with the PM, design solutions with the designer, build and ship them yourself, then watch the traces to know if it works. The loop is tight and you own all of it.

AI at the core of a 15‑year‑old profitable marketplace. This isn't a startup burning cash on an AI bet. It's an established travel company with real revenue, real users across 160+ destinations in over 5 markets, and a strategic conviction that AI will reshape its core experience. Startup autonomy with business‑model gravity.

What's in it for you?

La Voyagerie: our lively HQ in the 9th arrondissement of Paris

Based in Paris? Up to 3 days of remote work per week.

Fully remote? Up to 2 consecutive days on‑site per month when the whole team comes together in Paris.

Travel budget through the CSE to experience Evaneos trips.

Remote work from abroad several times a year.

25 paid leave days + 14 RTT days.

Choice of PC or MacBook Pro.

Wellpass (yoga, meditation, sports partially covered).

Cultural lunches, seminars, afterworks, and regular team moments.

Salary: €67–75K gross annual, based on experience.

What we're looking for

You may be a great fit for this position if:

• You have at least 5 years of software development experience and are comfortable with APIs and data stores.
• You’ve built and deployed AI agents in production with real users: not just PoCs or hackathon projects.
• You’ve set up or contributed to LLM evaluation pipelines (offline benchmarks and/or online metrics) and have designed guardrails.
• You have hands‑on experience with observability for LLM systems: tracing, monitoring, diagnosing regressions from traces (Langfuse, LangSmith, or equivalent).

Mindset & ways of working

• Curious and passionate about AI: you keep up with the state of the art, test things to understand them, and apply what you learn to what you build.
• Full‑stack at heart: in a small team, you’re happy to touch front, back, and infra. No “That’s not my job.”
• Strong ownership: you spot problems, propose solutions, and move forward while keeping everyone in the loop. You finish what you start.
• Product appetite: you want to understand users, conversion, and the business. Engineering is a means to an end: crafting meaningful experiences and solving real customer problems.
• Team player: you like pairing, sharing knowledge, communicating clearly, and pulling in the same direction.
• Fluent in professional English.

You’ll thrive here if:

• You want to build and ship AI products to real users, not do pure ML research.
• You care about the product outcome, not just the technical pipeline.
• You’re comfortable working daily alongside product, design, and business stakeholders.

We are committed to creating an environment where everyone feels respected and included. We strive to provide equal employment opportunities, regardless of gender, sexual orientation, background, disability, or any other characteristic that makes you who you are. Our goal is simple: to ensure that everyone can thrive, feel safe, and be treated with respect.

In French and/or English:

RH call with our Talent Acquisition Manager (Greismar): initial chat to get to know you, learn about your work experience, and how they connect to the role. 30–45 minutes.

Take‑home assignment (with your favorite coding agents): build an agent integrated into a travel product. We’ll share the brief, evaluation criteria, and expected deliverables upfront.

Technical interview with Romain Q and one engineer: ways of working, collaboration and deep‑diving into your assignment + discussion on agentic architecture and system design. 1h30.

Product discussion with Lou or Clément & Jocelyne: discussing how you work with product teams and make informed decision. 1h.

Final conversation with Romain R, Chief Product Officer. 30 minutes.

A quick word on how we read your application. To give every candidate a fair shot and the attention they deserve, we sometimes lean on AI tools to support parts of our hiring process: to help review applications or assess responses. These tools support our team and help us stay consistent, but they don’t make decisions for us. Every final hiring call is made by a person. Questions about how we process your data? Feel free to reach out.