Sina Barazandeh
I am a PhD student in the School of Computer Science at Carnegie Mellon University (Computational Biology Department), advised by Dr. Jose Lugo-Martinez.
I build machine-learning systems for agentic scientific discovery and autonomous laboratories. My work combines AI agents, reinforcement learning, generative models, and ML systems to plan experiments, operate physical equipment safely, recover from failures, and learn from experimental results.
Most recently, I led CMU’s work using the Model Hardware Standard (MHS) in the limited research preview: I built instrument drivers and orchestration so an AI agent could coordinate a liquid handler, plate reader, robotic arm, and cameras for closed-loop dose-response experiments.
I am seeking Summer 2027 research and applied-science internships on teams building and deploying ML systems, primarily for agentic scientific discovery, autonomous laboratories, lab automation, and AI for science, and also in generative AI, reinforcement learning, and ML systems.
[Resume (PDF)] · [Academic CV (PDF)]
Education
- Ph.D. Computational Biology — School of Computer Science, Carnegie Mellon University (2024–2029)
Advisor: Dr. Jose Lugo-Martinez - M.Sc. Computer Engineering — Bilkent University (2021–2024, GPA: 3.9/4.0)
Advisor: Dr. A. Ercument Cicek · Thesis: Generative Models for Generating and Optimizing Biological Sequences - B.Sc. Computer Engineering — Shiraz University (2016–2021, GPA: 3.5/4.0)
Research Highlights
- Autonomous Laboratories / MHS: Led CMU’s work using MHS in the research preview; built drivers and orchestration for a liquid handler, plate reader, robotic arm, and cameras
- Agentic Scientific Discovery: AI agents for closed-loop experiments, autonomous orchestration, and fault recovery
- Simulation & Benchmarking: simPCL — simulator and benchmark generator for programmable cloud laboratories
- Generative Models: Sequence design with GANs and seq2seq models (UTRGAN, RNAtranslator)
- Reinforcement Learning: Policy optimization for biomarker discovery under structure-aware rewards
- Representation Learning: Evolution-informed protein subgraph features for downstream prediction
- ML Systems: FlashAttention and PagedAttention in MiniTorch (LLM Systems coursework)
Recent News
- Aug 2026: We used the Model Hardware Standard (MHS) in the research preview (announced by Anthropic) so an AI agent could operate a liquid handler, plate reader, robotic arm, and cameras. Serial-dilution dose-response experiments ran ~3× faster. I led this work at CMU. Read the post. Also covered by CMU SCS on LinkedIn and X.
- Contributed talk at the AI Scientist Summer Workshop, Microsoft Research, Cambridge, MA (Aug 2026): Agentic Fault Recovery for Autonomous Orchestration of Programmable Cloud Laboratories
- Attended SciFM 2026 (Chicago, IL)
- Two publications in 2025 (PLOS Computational Biology, Bioinformatics Advances) and RNAGEN in ChemBioChem (2026)
- Attended the SenNet Annual Meeting (Washington, DC, 2025) and PathVisions (San Diego, 2025)
See my projects, publications, and CV for more details.