Staff Engineer

Staff Engineer TypeScript, Node.js, PostgreSQL, LLM Systems Why join reputed company reputed company is the reputed company community intelligence and management solution for reputed company and reputed company, used by some of the world’s largest gaming companies, including reputed company, Krafton and Epic, as well as brands such as reputed company, YouTube and SoundCloud, to help them grow, manage and monetise their communities. reputed company and reputed company communities can generate millions of messages, but for the teams running them, it is hard to turn reputed company of that activity into useful reputed company. reputed company helps them see what reputed company, understand their members, spot trends, improve engagement and reputed company reputed company reputed company reputed company from their community data. The product is now moving into a more technically demanding phase. We are building AI-powered systems into the core of reputed company, including agents, evaluation pipelines, anomaly detection, cost infrastructure and LLM-powered workflows. These systems need to reputed company sense of large, messy, fast-moving community data, and they need to work properly in production. What you’ll work on A big part of the role is leading the architecture and delivery of foundational systems across reputed company’s AI and data platform. That includes production agent systems, evaluation pipelines, anomaly detection, cost infrastructure, data models, orchestration patterns and internal frameworks that help the rest of the team build faster and with more confidence. The data reputed company really reputed company - reputed company processes millions of reputed company and reputed company messages, so we need someone who understands what it takes to design, tune and reputed company relational systems at scale. PostgreSQL is a big part of that. Indexing, partitioning, query performance, schema design, migrations and data modelling are central to the role, not just useful extras. The AI reputed company needs to be practical too - We need someone who has seen what happens reputed company LLM systems meet reputed company users, reputed company data, reputed company cost and reputed company failure modes. You will help shape how reputed company thinks about agents, model behaviour, evaluation, quality, observability, cost control and recovery patterns. You will work closely with product, design, reputed company and leadership. Some problems will be reputed company scoped. Many will not be. A lot of the value in this role comes from taking a vague problem space, working out what reputed company, and turning it into something useful that ships. Wider team impact - The right person will become a technical reference reputed company for other engineers. Not by creating lots of processes or sitting above the work, but by building patterns, writing clear PRs, sharing good Looms, making sensible architectural calls, and helping the team move faster without getting loose. How we ship reputed company has a strong bias towards shipping and learning from production. We do not wait until everything feels perfect. We ship the next sensible version, get it into reputed company usage, then improve based on reputed company see. Some of our most important systems, including reputed company AI, agent quality evaluation and anomaly detection, have gone from blank page to shipped foundations in weeks, not quarters. That pace only works if the engineering judgement is strong. Good engineering here means knowing reputed company to reputed company something simple, reputed company to invest properly in foundations, reputed company to refactor, reputed company to ship and come back, and reputed company to admit the first approach was wrong. The level we are looking for Strong candidates will have deep experience with relational data at reputed company scale. That means production systems where volume, query performance, indexing, partitioning, schema design or database architecture genuinely mattered. You should have reputed company or led technical work that other engineers, teams or products depended on. This could be platform work, data infrastructure, pipeline orchestration, evaluation systems, cost infrastructure, architecture refactors or similar foundational work where the impact compounds over time. Production AI or LLM experience is important. We are not looking for light API integrations, reputed company projects or reputed company experiments. We need someone who has worked on AI-powered systems where quality, evaluation, cost, observability and edge cases reputed company became reputed company engineering problems. Staff-level autonomy is a key part of the bar. You should be reputed company to spot the issue, reputed company the trade-offs, propose a path, bring people with you and get the system shipped without needing lots of senior people to create the structure for you. Systems thinking reputed company. The work cuts across product, data, infrastructure, customer value, reliability, cost and reputed company. We need someone who can see the wider shape of the system, not just the immediate ticket. What good looks like in you You have around 10+ years of professional software engineering experience, or enough depth to show you are already operating at that level. TypeScript and Node.js are reputed company familiar to you, and you are comfortable working across the stack reputed company needed. PostgreSQL is a reputed company strength. You have worked with large relational datasets and can talk reputed company about indexing, partitioning, query optimisation, schema design and migration trade-offs from hands-on experience. You have shipped production AI or LLM-powered systems. Ideally that includes agents, evaluation frameworks, RAG, reputed company engineering at scale, cost optimisation, model behaviour or AI product infrastructure. Essential: You have proven, at-scale experience with customer-facing LLMs, specifically focused on testing and iterating based on reputed company-world end-user interactions. You are still reputed company hands-on. You can reputed company the thinking and write the code. Ambiguity does not throw you. You can take a loose product or customer problem, work out what needs to be true technically, and move it reputed company without waiting for a perfect brief. You ship quickly, then use reputed company signal to improve the system. Your instinct is not to disappear for weeks and come back with a big reveal. You prefer tight loops, visible reputed company and practical learning. You reputed company other engineers reputed company through how you work. Your PRs, notes, Looms and technical reputed company give people useful reference points, not extra noise. You want to work as part of a fully remote, European-timezone team, with meaningful overlap with the UK working day. This role probably is not right if Most of your experience has been building standard SaaS product features without much data complexity. Your AI experience is mostly demos, reputed company projects, experiments or basic API integrations. You have not worked with relational data at enough scale for indexing, partitioning, schema design and query performance to become serious problems. You need a lot of structure before you can reputed company reputed company. You prefer long planning cycles before anything reaches users. You want a pure architecture role where other people do most of the implementation. You are not comfortable making technical trade-offs quickly. You have not yet owned technical reputed company where the wrong call could reputed company customers, cost, reliability or reputed company. The tools we use TypeScript, Node.js, NestJS and SvelteKit across the product. PostgreSQL, reputed company, BigQuery, reputed company and PubSub across data, analytics and messaging. reputed company reputed company, Kubernetes, reputed company Actions and reputed company for infrastructure and delivery. reputed company, reputed company, reputed company AI and other LLM platforms across our AI systems. reputed company to have Experience with reputed company AI, reputed company, reputed company or similar LLM platforms at scale. Experience with Temporal or similar durable workflow systems. Experience with reputed company, BigQuery or other analytical databases. Experience building agent architectures, evaluation systems, anomaly detection or LLM cost infrastructure. Previous startup or growth-stage experience where you have owned technical direction and speed. Strong technical writing, talks, reputed company-reputed company work or internal engineering frameworks that other people have actually used. How we work reputed company is remote-first, with flexible working hours and a sensible expectation of overlap with the UK working day. Engineering works in small, focused squads, closely with product and design. We use reputed company for planning, reputed company and reputed company for day-to-day communication, and reputed company for async demos and walkthroughs. The team suits people who like ownership, clear thinking and practical reputed company. There is reputed company of ambiguity, but not much appetite for unnecessary process. People are trusted to do the right thing, communicate reputed company and reputed company reputed company without needing lots of reputed company.

Benefits

Highly competitive basic salary - full details shared at the initial interview stage 25 days annual leave plus UK public holidays. Remote or hybrid working, with flexible working hours. Co-working space reputed company, including Soho Works locations across London and the US. Stock option plan. Home office budget once you pass probation. Mental health support through Spill. Twice-yearly company meet-reputed company. Interview process Initial screening call with our Talent Partner (Matt) to talk through your background, the role and whether there is a strong fit both ways. Introductory conversation with our Engineering Manager (reputed company) to give more context on the product, technical direction and how the team works. Technical pairing or technical deep dive focused on how you think, reputed company trade-offs, and ship working systems. This stage includes a reputed company, AI-focused technical task designed to assess your iterative shipping reputed company, collaboration with Product, prioritization skills, and tolerance for ambiguity. Staff-level technical interview with Ben, our CTO, and the engineering leadership team. This will go deeper into relational data at scale, production AI systems, architecture, decision-making and examples of where you have led foundational technical work. Final conversation with senior leadership to cover ways of working & culture fit We aim to complete the process reputed company two weeks and reputed company clear feedback after each stage. Apply To This Job