Software Engineer (AI specialist)

We live in a world that runs on physical industries like energy, logistics, manufacturing. These sectors are essential to our economy and daily lives, yet they are held back by inefficiencies, fragile legacy systems, and decades-old ways of working. This is not just a business problem; it is a societal one. When grids fail, supply chains stall, or factories falter, entire economies are disrupted.

Overview

We are looking for a creative Software Engineer with a passion for developing scalable, reliable, efficient, and beautiful AI‑first software and a strong sense of ingenuity. You will be part of an all‑star early team, building the future of software. We have been building product with AI at its core for enterprise customers since the early days of GPT‑3.5 to the latest and greatest models and have built internal tooling to work efficiently with LLMs. You will help develop frameworks for evaluation, verification, specialized code and test generation pipelines, agentic behavior, RAG, observability and orchestration tooling that are used across our product and inform real customer needs.

What You\’ll Do

• Define, build, and maintain reliable AI‑first product features, including internal frameworks and tooling for working with LLMs
• Experiment with various AI/LLM techniques, models, and approaches to increase the accuracy and reliability of our outputs
• Build and improve evaluation systems and benchmarks
• Design and build reusable application architectures and toolchains, working closely with the software synthesis team
• Write and ship high‑quality, efficient, and reliable code that is extendible, scalable, and observable – balancing experimentation with shipping
• Take ideas from experimentation to shipping, ensuring a great product experience
• Build and help foster an engineering culture that encourages innovation and high‑quality code standards while shipping fast
• Collaborate closely in a small, highly motivated team to build world‑class software

What We\’re Looking For

• At least 3 years of relevant work experience
• The drive to work at an early‑stage startup in a fast‑paced, quickly changing, and independent environment
• Proficiency in modern typed Python, Docker, and PostgreSQL
• Experience in a compiled typed language like Go or Rust, or modern front‑end development (plus)
• Deep familiarity with modern Python ecosystems and toolchains, and experience working with large codebases
• Creative experimentalist at heart, with industry experience where experimentation was essential
• Professional experience using LLMs or enthusiastic experiments in your own time; industry experience with LLM toolchains (e.g., LangChain, observability tooling, RAG systems) is plus
• A fundamental understanding of code analysis, compilers, domain‑specific languages, or related topics, and a keenness to learn in these areas
• A passion for clean and reusable interfaces, with the ability to iterate fast
• Capability to work closely with a product manager to turn customer requirements into reliable systems quickly, balancing long‑term maintenance; excellent communication of technical systems to product‑focused stakeholders
• Familiarity with e2e testing frameworks, large language models, CI/CD tooling, workflow orchestration systems such as Argo or Temporal, and/or experience with Kubernetes (plus)

Special note about career breaks: We accept candidates with career breaks and CV gaps. Life happens, we get it. Please apply anyway!

What We Offer

• Competitive salary and share option plan
• 25 days annual leave, plus public holidays
• Excellent pension scheme, matched to 7.5%
• Private healthcare, including family cover and pre‑existing conditions
• Flexible hybrid working with 2 days per fortnight in our Farringdon office, with schedules negotiable
• Regular social events and offsites
• Access to a gym and fitness classes in our office building

Seniority Level

• Mid‑Senior level

Employment Type

• Full‑time

Job Function

• IT Services and IT Consulting

Referrals increase your chances of interviewing at Cogna by 2x