AI Architect

Quantomatic is a small quantitative proprietary trading team. We are building a system that lets researchers delegate repeatable parts of their work to AI agents while keeping the evidence behind each result inspectable.

You will own the technical architecture and contribute directly to the implementation. Working with quantitative researchers, you will turn research tasks into reliable workflows and establish how the system is evaluated, monitored and improved.

• Define the interfaces between agents, datasets, research code, backtesting services and experiment records.
• Build workflows that prepare data, develop candidate experiments, execute tests and compare the resulting evidence.
• Give agents well-defined tools, execution permissions and recoverable state; make failures visible and expensive runs controllable.
• Preserve the data versions, code, parameters and outputs needed to trace and reproduce each experiment.
• Develop evaluations for numerical correctness, task completion and research usefulness, using deterministic checks where possible.
• Work with quants to account for look-ahead bias, repeated testing and realistic trading constraints as experimentation scales.
• Measure reliability, cost and researcher time saved. Define clear review points before research outputs enter production trading.

• Experience owning the architecture and delivery of substantial AI, data or research systems, with strong hands-on software engineering skills.
• Strong Python and system design skills, including APIs, data pipelines, job execution and observability.
• Practical experience with LLM tool use, agent workflows and evaluation beyond an initial demonstration.
• The ability to distinguish a plausible explanation from a result supported by data and computation.
• Enough statistical and quantitative research fluency to design experiments and evaluation methods with researchers.
• Sound judgement about when to use an agent, a deterministic program or a human decision.

Systematic trading research, market data, simulation or backtesting infrastructure would be particularly relevant. So would experience building experiment platforms or operating AI systems with strict data-access and compute constraints.

This role can be based remotely or in our Bangkok office. You will work directly with the research team and take responsibility for architectural decisions, implementation quality and how well the system supports real research.