Sergei A. Nikolenko

Staff scientific ML engineer and computational chemistry lead building practical AI systems for molecular discovery.

My work spans molecular docking, protein-ligand modeling, generative molecular design, ADMET/QSAR, chemical LLMs, spectra workflows, scientific benchmarks, and reproducible Python/HPC infrastructure.

Portfolio · LinkedIn · ORCID · Kaggle

Selected public work

• Matcha — multi-stage Riemannian flow matching for physically valid molecular docking.
• HEDGEHOG — stage-based evaluation for generative molecular design.
• Bento — reproducible protein-ligand docking benchmark and HPC workflows.
• SynthLadder — chemistry benchmark and evaluation package for agentic LLMs.
• posecheck-fast — high-throughput docking pose validation.
• Burrete — macOS molecular previews and Quick Look tooling.

Focus

• AI-driven drug discovery and molecular modeling
• Scientific ML benchmarks, reproducible pipelines, and research infrastructure
• Chemical LLMs, tool-enabled evaluation, and structured scientific workflows
• Product-grade research tooling for computational chemistry teams

Collaboration

Open to research and engineering collaborations in computational chemistry, molecular modeling, drug discovery AI, chemical LLMs, molecular spectra, and scientific software infrastructure.