SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Coupa Software is seeking a Senior Data Scientist to join the Data Capture Research team in Prague, focusing on sourcing optimization within their total spend management platform. The role sits at the intersection of multiple ML disciplines applied to a greenfield problem space: how procurement decisions are made, from supplier discovery and matching across a 10M-node network, to should-cost modeling, auction design, award allocation under constraints, and document understanding of RFPs, specs, quotes, and contracts.
You will own sourcing research initiatives end-to-end, from problem framing on real spend data through experimentation to production deployment. The work spans recommendation and retrieval systems, forecasting and should-cost modeling, game theory and mechanism design, combinatorial optimization, and multimodal document understanding leveraging Coupa's proprietary T-LLM (transactional LLM) architectures trained in-house.
Key responsibilities include: defining datasets, labels, and benchmarks that make sourcing problems learnable; designing and running experiments with fast prototypes and honest baselines; building on T-LLM foundations for document-based signals; shipping with deployment constraints in mind (inference cost, latency, robustness); and collaborating across Sourcing product, data, and AI Platform teams.
You will work in a small, senior team of researchers and engineers—the group that built Rossum's production models from scratch—with direct access to Product and the AI Platform team. Ideas that work roll into systems used at scale by companies whose margins depend on sourcing decisions.
Coupa's sourcing platform operates on $10 trillion of transacted spend data across 10M+ buyers and suppliers globally. Modern ML has barely been applied to sourcing; this is genuinely greenfield territory with real ownership and a short path to customers.
REQUIREMENTS:
- 5+ years in applied ML, data science, ML engineering, or quantitative research, with models or decision systems taken into production
- Strong Python and comfort with messy, large-scale data (SQL, data quality and trustworthiness)
- Depth in at least one modeling discipline with curiosity about others: deep learning, recommendation/ranking, forecasting/time series, optimization/operations research, causal inference/econometrics, RL/bandits, market/mechanism design, or LLM-based systems
- Scientific rigor: strong experiment design, healthy skepticism about metrics, care about leakage, baselines, and evaluation that survives production contact
- Ownership and comfort in greenfield, ambiguous problem spaces; ability to engage with product and domain experts
- Interest in procurement, supply chains, or market design welcome but not required
