Research Engineer

About Us

Artificial Societies (societies.ai) helps Fortune 500 organisations understand how real-world audiences think, feel, and respond — without the cost, delay, or limitations of traditional research. We build large-scale simulated populations grounded in real-world data, enabling clients to test messages, strategies, and concepts. Our clients love the ability to receive insights backed by millions of responses in hours rather than months, and to access previously inaccessible audiences (e.g. investors, CEOs, opinion leaders).

We're backed by Y Combinator, Point72 Ventures, and Kindred Capital, plus investors from DeepMind and Sequoia Scout. We work with leading organisations across technology, financial services, communications, and public affairs.

We're a lean and highly effective team. We move fast, hold a high bar, and trust each other to own outcomes end-to-end. The culture is collaborative but low-ego: ideas win on merit, not seniority.

The Role

We're looking for a Research Engineer to advance the frontier of our simulation engine: the science and systems that make a simulated population statistically faithful to a real one. You'll ground our personas in real-world data, post-train models, and design the evaluation methodology that tells us how well our simulations match reality. You'll work directly with our CEO and Head of Research, and partner closely with our agent, product, and forward-deployed engineers to move ideas from a research notebook into production code our enterprise clients can rely on.

This is a role for someone who lives at the intersection of research and engineering. You form a hypothesis, build the infrastructure, run the experiment, analyse the result, and decide what to try next. You care as much about a clean experiment and a well-calibrated metric as you do about code that runs reliably at scale. Our accuracy benchmarks are grounded in behavioural science and psychometrics; you'll be the person who pushes them forward.

What You'll Do

• Advance the simulation engine. Improve how faithfully our populations reproduce real human opinion distributions and how coherently individual personas hold beliefs across long interactions.
• Ground personas in real-world data. Build the pipelines and methods that turn survey, behavioural, and audience data into simulated populations that reason like the people they represent.
• Train large-scale models. Apply frontier ML methods in supervised learning and natural language processing to develop state-of-the-art simulations of individuals, subgroups, and populations.
• Own evaluation and calibration methodology. Design the evals, benchmarks, and validation studies that measure simulation fidelity against real-world decisions, and make the results interpretable to the team.
• Run experiments end-to-end. Hypothesis, infrastructure, training run, analysis, decision. Build the experimentation infrastructure that lets us iterate fast and trust the numbers.
• Bridge research and production. Partner with agent and product engineers to turn validated research into capabilities that ship, and with FDEs to stand behind results for high-stakes client work.
• Raise the scientific bar. Bring rigour, honesty about uncertainty, and taste to everything from dataset construction to metric design.

What We're Looking For

• At least 2+ years in a research engineering, ML, or applied research role, or a research background (e.g. publications, a relevant graduate degree) with strong applied engineering skills. You've taken research ideas to working systems.
• Python-strong. Our simulation engine is Python-based and this role lives in it day to day.
• Hands-on with the ML research loop. Designing, running, and analysing experiments, and drawing concrete conclusions from noisy data.
• Experience with post-training and evaluating LLMs. RL, dataset and reward design, synthetic data, or related training methods.
• Statistically fluent. You reason carefully about distributions, calibration, and validation, and you care whether your metrics measure what you think.
• Comfortable building reliable data pipelines and experimentation infrastructure.
• Ownership mentality. You don't wait for a ticket. You see what needs to happen, and you make it happen end-to-end.

Nice to Have

• Background in behavioural science, computational social science, psychometrics, statistics, or a related empirical field.
• Publications or open-source work in LLM simulation, evaluation, calibration, or human-behaviour modelling.
• Experience with distributed training or scaling ML systems.
• Familiarity with TypeScript (our product stack) to collaborate smoothly with product engineers.
• Experience scaling a company from seed to later stages.

Logistics

• Location: London, in-office. We build better together, and this role is no exception.
• Salary: £80,000 – £150,000 depending on experience.
• Equity: Up to 0.20%.

Why Join Now

This is a rare chance to define the science behind a genuinely new technology that lets the world's most important organisations test ideas on simulated populations before they touch the real world. You'll work directly with our CEO and Head of Research on our simulation engine. If you want the ownership of an early team, the craft of a product you're proud of, and the technical depth to build it properly, this is the role.