Why this opportunity stands out
Senior AI Engineer for a UK applied-AI lab. Multimodal generative systems, enterprise agents, media and content pipelines. £100k–£150k, hybrid, ASAP start.
About the role
## What you will own
You will be a senior engineer inside an established applied research and engineering lab. Small team with big autonomy, already shipping. Your mandate is to take generative AI out of the demo and into production-grade systems used by media, content and enterprise teams every day.
You will design, build and operate the agentic and generative backbone behind our partners' most ambitious products:
- **Enterprise agents** that orchestrate tools, knowledge and people across multi-step business workflows. Procurement, content production, compliance, customer operations and internal research.
- **Multimodal systems** that reason over text, structured data, audio, image and video. Not just chat, but document understanding, visual content analysis, voice-driven interfaces and video comprehension pipelines.
- **Content and media pipelines** that turn generative models into a real authoring and editorial surface for newsrooms, content studios, marketing teams and learning platforms.
- **The platform under all of it**. Evaluation, observability, guardrails, evals, cost control, prompt and context engineering as a first-class engineering discipline.
## What you will have done before
- **Five or more years building production software**, with at least the last 2 to 3 years focused on generative AI in earnest: LLMs, multimodal models, agent frameworks, retrieval and tool use.
- Shipped at least one **agentic system** to real users. Something that calls tools, plans, retrieves, iterates and recovers from failure, not a single-turn prompt.
- Built or significantly contributed to a **multimodal pipeline**. Image, audio, video, OCR, speech-to-text, embeddings across modalities, or a creative-tooling system that handles non-text content.
- Have an opinion on **how to evaluate generative systems**, and have built eval harnesses rather than just hand-waved.
- Comfortable across the **modern Python AI stack**: PyTorch or JAX, Hugging Face, vector stores, orchestrators such as LangGraph, CrewAI or AutoGen, plus the usual cloud and data infrastructure underneath.
- **Media or content domain experience** is a strong plus. Newsrooms, studios, publishing, learning, marketing platforms, anywhere generative output meets editorial judgement and brand standards.
- You write code other engineers actually want to read, and you can explain a complex model decision to a product or commercial stakeholder without losing either of them.
## What success looks like in the first six months
- A flagship **multimodal agent** is live with a major partner, handling real volume and improving week over week.
- The team has a **reusable agent platform**. Orchestration, retrieval, evaluation, observability and cost control, built so the next agent is ten times faster to ship on.
- You have **raised the bar** through code reviews, design docs, hiring and mentoring, making the lab measurably better at shipping generative AI safely and quickly.
## Compensation
- **£100,000 – £150,000** base, depending on experience and depth.
- Significant equity and share scheme in the lab.
- Generous learning and conference budget, plus dedicated research time.
- Real hybrid working, typically two to three days a week in lab space in **London, Manchester or Birmingham**, with the rest from wherever you do your best thinking.
## Location and start
- Based in the UK, **hybrid** out of London, Manchester or Birmingham.
- **ASAP start**. The lab is shipping now and the next three quarters are already overcommitted.
Key responsibilities
- Build the platform layer underneath: orchestration (LangGraph / CrewAI / AutoGen-class), retrieval, evaluation, observability, cost control and guardrails.
- Partner closely with research, product and commercial leads to translate ambitious briefs into shippable scope, and to push back when they should.
- Set the technical bar for generative AI engineering across the lab: code quality, design docs, reviews and hiring.
- Mentor mid-level engineers and help grow the team.
What they're looking for
- 2+ years focused on generative AI: LLMs, multimodal models and agent frameworks.
- Production experience with at least one of: PyTorch, JAX, Hugging Face Transformers / Diffusers, large-scale inference.
- Hands-on with vector stores, RAG, prompt and context engineering, tool-use agents.
- Strong Python; comfortable across the modern data and infra stack.
- Media / content / publishing / studio experience is highly valued.
- Right to work in the UK; based in or willing to relocate to London, Manchester or Birmingham.
Benefits
- Real hybrid working from London, Manchester or Birmingham labs.
- Dedicated research time, conference and learning budget.
- Work on problems that are genuinely at the edge of what multimodal generative AI can do in production.