Founding AI Research Engineer (up to £140k)

Dex; Greater London; On-site; GBP 70,000 - 120,000; Full time

Job summary

Dex is partnering with a company building a proactive AI assistant that maintains a living model of a person’s work history, and acts on their behalf. The team aims to handle long-term memory, retrieve messy context, and reason live on screen across diverse sources.

You will design retrieval systems, develop reasoning models, and own the end-to-end lifecycle from hypothesis to production, using Python and cloud infrastructure in a hands-on role based in Elephant and Castle, five days a week.

Qualifications

• PhD or equivalent end-to-end technical project at a company.
• 5+ years solving research-level NLP/LLM problems in academia or industry.
• Implement and build upon current research papers, not just apply libraries.
• Experience forming/testing hypotheses and building robust, scalable systems.

Responsibilities

• Design and implement novel retrieval and knowledge-base systems for a persistent, living model of user context.
• Develop and deploy reasoning models across diverse, messy data sources.
• Build models capable of real-time reasoning over live screen and application context.
• Take research from hypothesis to production, owning the lifecycle.

Skills

Python

NLP/LLMs

Cloud infra

SQL/NoSQL

Research & prototyping

Education

PhD in CS/ML

Tools

PyTorch

Docker

Job description

This role is with one of Dex's trusted partner companies. We work closely with their teams to truly understand their culture, goals, and what they're looking for, so we can match you with the right opportunity and give you context about the role before you commit to a process.

The role

This company is building a proactive AI assistant that maintains a living model of a person's entire work history — emails, messages, meetings, documents. It acts on their behalf, drafting replies and preparing for meetings, all before being asked. The core challenge is building long-term memory that stays current, retrieving messy context across diverse sources, and reasoning about information live on screen. This is a genuinely unsolved problem in AI.

The work

• Design and implement novel retrieval and knowledge‑base systems for a persistent, living model of user context.
• Develop and deploy reasoning models that operate across diverse, messy data sources.
• Build models capable of real‑time reasoning over live screen and application context.
• Take research from initial hypothesis and experimentation through to robust, production‑ready systems, owning the entire lifecycle.

What You Bring

• A PhD or personally owned a hard, end‑to‑end technical project at a company.
• 5+ years solving research‑level problems in NLP, LLMs, or adjacent ML, in academia or industry.
• You implement and build upon current research papers, not just apply existing libraries.
• You've run both halves of the work: forming and testing hypotheses, and building robust, scalable systems.
• Strong Python, comfortable with cloud infrastructure, and both SQL and NoSQL databases.
• Work in person in Elephant and Castle five days a week.

As part of the recruitment process at Dex, we process your personal data in accordance with our Privacy Notice for Job Applicants. This notice explains how and why your data is collected and used, and how you can contact us if you have any concerns.