Mohamed A M Elansary, PhD
Target: Research Scientist, Data — Periodic Labs
Sourced insights
- Evaluations and data + tight feedback loop: Most important aspect of Scientific AI creation is evaluations and data — cutting-edge evals from scientific use cases, external datasets, experimental data into the training stack, RL environments; goal is a tight feedback loop between scientific use cases, model evaluation, and training data. Source: Ashby JD — Research Scientist, Data · mirror Built In
- AI scientists + autonomous labs: Periodic is building AI scientists and the autonomous laboratories for them to operate — labs supply GBs of high-quality data and negative results, and give AI scientists tools to act; in the physical sciences, nature is the RL environment. Source: periodic.com
Proof — evals + data strategy × scientific datasets × pipelines × provenance × UQ × ship
- PhD Environmental Engineering, TAMUK 2022: multimodel / ensemble surface-water/groundwater forecast uncertainty quantification & reduction on HPC — fair baselines, imperfect ground truth, skill before claims.
- Scientific data stacks: USGS/NOAA/NASA multi-source QA with provenance discipline — same judgment shape as sourcing chemistry/physics/materials/sim/lab datasets for training + eval.
- Pipelines + tooling: Vertexium production agentic LLM + retrieval + multi-tenant agents (inspect → synthesize); Lucent CTO monitoring; automated env-monitoring → validated reporting.
- Physical-science PhD + research mindset: hypothesis → controlled multimodel comparison → iterate; AMS multimodel streamflow; AMS 2021 floods & droughts.
- Honest frame: no Periodic lab ownership claimed; no materials/superconductor claims. The PhD who ships. Menlo Park / SF onsite OK · Prefer take-home.