Senior ML/AI Engineer
About Depict
Depict.ai is redefining the future of fashion e-commerce with AI. Founded by Oliver Edholm and Anton Osika, we’ve entered a new growth phase with products that our customers truly love. From AI-powered merchandising to GPT-driven search, we’re pioneering solutions that make e-commerce smarter, faster, and more personalized. Our customers span the Nordics, UK, USA and now, we’re looking for exceptional engineers to help us shape the next chapter.
The role:
We’re on the hunt for a Senior ML/AI Engineer to help us bring this vision to life. You'll be joining a lean, high-impact team with full ownership over everything from system architecture and modeling to infrastructure.
If you're excited about working with LLMs, product search, and building innovative solutions from the ground up, we’d love to hear from you.
We collaborate with some of the fastest-growing DTC brands and nine-figure online retailers across Europe and the U.S. Depict is backed by world-class investors including Y Combinator, Initialized, EQT Ventures, and Northzone.
What You’ll Do
- Support the development of LLM-powered conversational search—tuned for ecommerce catalogs, user behavior, and brand voice.
- Build advanced systems that combine RAG, semantic search, and multi-turn dialogue to handle real-world ambiguity.
- Architect pipelines that blend structured product data, reviews, and customer signals into compelling search experiences.
- Prototype with models like GPT, Mistral, LLaMA, or fine-tuned domain-specific models.
- Development across the full ML lifecycle—from training and embeddings infra to scalable inference.
- Collaborate with product and design to craft UX that feels magical, not generic.
- Define and run evaluation strategies—from offline metrics (e.g., Recall@K, nDCG) and human-in-the-loop feedback to online A/B testing tied to engagement and conversion.
- Build tooling to ingest, clean, and version product catalogs, reviews, and chat logs.
- Stay on top of research, and make smart calls on what to build vs. adopt.
- Ship fast, learn from usage, and balance startup speed with technical excellence.
Who You Are
- You’ve built and shipped production ML systems, especially in search, NLP, or conversational AI.
- You’re skilled in transformers, LLMs, vector search, and semantic retrieval.
- You’ve designed evaluation frameworks and know how metrics tie back to real user value.
- You work across the stack—modeling, APIs, data pipelines, infra, and even frontend.
- You thrive in early-stage environments, moving fast and wearing many hats.
- You care deeply about user experience, not just model accuracy.
Bonus Points
- Experience with multi-turn conversations, chat UIs, or personalization in search.
- Experience with fine-tuning, building or deploying machine learning models
- Familiarity with OpenAI, LLaMA, Mistral, or embedding models like E5 or Cohere.
- Worked on model safety, hallucination detection, or guardrails for generative apps.
- Exposure to quantization, PEFT, or retrieval tuning.
- You've measured success using both quantitative metrics and qualitative user feedback.
- Familiarity with safety/guardrails in generative applications.
Why You’ll Love Working Here:
- Competitive salary and meaningful equity opportunities.
- 30 days paid vacation + wellness and learning allowances.
- Work on the bleeding edge of AI in e-commerce.
- Collaborate with a tight-knit, ambitious team in Stockholm
- See the impact of your work on real customers around the world
- Department
- Engineering & Product
- Locations
- Stockholm
- Remote status
- Hybrid
Colleagues
Stockholm
This is Depict.
National programming champions, art school dropouts, CERN physicists and former professional athletes. An interdisciplinary team brought together by a common goal.
Are you one of us?
About Depict
Depict is the best merchandising solution for fashion E-commerce. We make merchandising simple, helping you create standout shopping experiences to your customers in real-time, with smarter workflows.
Already working at Depict?
Let’s recruit together and find your next colleague.