AI Engineer

Plainsight; Antwerpen; Sur place; EUR 60 000 - 80 000; Plein temps

Avantages offerts par ce poste

Real growth plan

Lunch conversations on diverse topics

Competitive pay

Résumé du poste

Plainsight in Antwerpen is looking for someone to build AI that works reliably in production environments. You will work on intelligent assistants and AI products, engaging with Python and machine learning models.

The ideal candidate has strong Python skills, hands-on experience with generative AI, and good judgment about AI model deployment and risks. This role offers competitive pay and an opportunity to be part of a dynamic team.

Qualifications

• Hands-on experience building or fine-tuning generative AI models.
• Ability to ship AI solutions to cloud environments without ceremony.
• Good judgment in managing data risks associated with AI.

Responsabilités

• Develop and deploy intelligent assistants and AI products.
• Design and manage RAG pipelines and vector databases.
• Collaborate with clients to ensure practical adoption of AI outputs.

Connaissances

Solid Python

Generative AI experience

Comfort with AI/ML stack

Architectural sense

Data privacy awareness

Outils

PyTorch

Hugging Face

LangChain

Description du poste

You build AI that earns its keep. Not the demo that dazzles a boardroom on Tuesday and quietly dies on Thursday. The kind that runs in production, helps real people get through real work, and holds up when nobody's watching it.

What you'll work on

Intelligent assistants, copilots, multi-agent systems, and AI products that slot into how people already work instead of asking them to change everything. Most days that means Python, the right model for the job (open-source like Llama or Mistral, or a commercial API from OpenAI, Anthropic, Azure OpenAI), and the unglamorous work of feeding it the data it actually needs to be useful.

The job is equal parts interesting and tedious: prompt design, RAG pipelines, vector databases, evaluations, fine-tuning when it's truly warranted (LoRA, PEFT), getting it deployed wherever it needs to live, monitoring, and the honest conversation afterwards about what to change.

What you bring

Solid Python. Hands‑on experience building or fine‑tuning generative AI: LLMs, RAG, agents, the occasional diffusion model. Comfort with the AI/ML stack of the moment (PyTorch, Hugging Face, LangChain, or whatever's replaced it by the time you read this) and the ability to ship it to a cloud without ceremony. Architectural sense. And good judgment about what should and shouldn't be handed to a model, plus the risks that ride along with it: data privacy, prompt injection, and the confident nonsense we politely call hallucination.

You don't need to tick every box. We did most of our learning on the job too.

How we work

We code as agentically as we can. Claude Code, Cursor, Copilot, whichever tool earns its place that week. If you already work this way, brilliant. If you don't yet, you'll pick it up here, because it's how the team ships.

Beyond the tools: we build with adoption in mind. The model is never the point. What matters is whether the people downstream trust the output enough to actually use it. So we pair with our clients' teams, show our work, and try hard not to leave behind something only we can keep alive.

What you'll get

A team that takes the work seriously and itself a lot less so. A real growth plan with your name on it. Lunch conversations that swing from data architecture to whatever strange thing someone read this week. Competitive pay.