Location: Eindhoven
Start Date: As soon as possible
Duration: 12 months
Contract Type: Freelancer (ZZP) ONLY!
On-site Requirement: Minimum 3 days per week
Role Overview
You will join an AI Competence Center focusing on generative and agentic AI systems optimized for on‑device deployment. You will work on LLMs, LMMs, and VLA models, improving performance on NPU‑based hardware, translating research into production-ready edge solutions.
Responsibilities
* Optimize LLMs and multimodal models for device deployment
* Apply quantization (8-bit, 4-bit, mixed precision), pruning, distillation
* Implement inference acceleration methods such as speculative decoding
* Enhance performance of small language models enabling tiny agents at the edge
* Deploy optimized models via Ollama, llama.cpp, ONNX Runtime, TFLite
* Create benchmarking pipelines for generative/agentic systems
* Build PoCs for industrial safety monitoring, in‑cabin sensing, etc.
* Translate advanced techniques into product-ready implementations
Profile Requirements
* MSc/PhD/EngD in a technical field
* 5+ years in AI/software engineering with LLMs, VLMs, performance systems
* Experience with quantization, pruning, optimization techniques
* Strong PyTorch / TensorFlow background
* Experience with agentic AI (LangChain, Google ADK, SmolAgents, etc.)
* Understanding of safety & guardrails for agentic systems
* Experience with deployment frameworks (CUDA, TensorRT, TFLite, ONNX, Ollama)
* Embedded systems/NPU accelerator experience
* Strong Linux, embedded architecture, version control, build systems
* ML‑Ops experience (MLFlow, ClearML)
* Knowledge of YOCTO / OpenEmbedded beneficial
* Excellent programming in C, C++, Python, Bash
* Excellent English communication skills
Let op: vacaturefraude
Helaas komt vacaturefraude steeds vaker voor. We waarschuwen je voor mogelijke misleiding:
* Wij zullen nooit via WhatsApp of in een videogesprek vragen om jouw persoonlijke gegevens (zoals een kopie van je ID, bankgegevens of BSN).
* Twijfel je over de echtheid van een vacature of contactpersoon? Neem dan altijd rechtstreeks contact met ons op via de officiële contactgegevens op onze website.
Important: job fraud
Unfortunately, job fraud is becoming more common. Beware of such scams:
* We will never ask for personal information (such as a copy of your ID, bank details, or social security number) via WhatsApp or during a video call.
* If you're unsure whether a vacancy or contact person is legitimate, please reach out to us directly using the official contact details on our website.
Key Skills
Ranked by relevance
Related Jobs
3 roles aligned with this opportunity
Forward Deployed Engineer, GenAI, Google Cloud
2026-05-20
Software Engineer III, Embedded Software, Fitbit Device
2026-05-26
Software Engineer III, Wear Core Platform
2026-05-20
- Posted
- Mar 11, 2026
- Type
- Full-time
- Level
- Entry
- Location
- Eindhoven
- Company
- Huxley
Industries
Categories
Related Jobs
3 roles aligned with this opportunity
Forward Deployed Engineer, GenAI, Google Cloud
2026-05-20
Software Engineer III, Embedded Software, Fitbit Device
2026-05-26
Software Engineer III, Wear Core Platform
2026-05-20