Mohamed A M Elansary, PhD
Target: Member of Technical Staff — Research Engineering, Evaluation — Causal Labs
Sourced insights (≤2)
- Measurement trustworthiness: Progress is only as trustworthy as its measurement — every team needs to know precisely and comparably whether a change made the model better. Seat builds the central evaluation framework (pipelines, metrics, tools → shared understanding). Source: Ashby JD — Research Engineering, Evaluation
- Weather as LPM training ground: Weather is the ideal training ground for a Large Physics Model — most well-observed physical system; rapid, objective ground-truth feedback from sensory observations at a scale that dwarfs today’s LLM training data. Source: causallabs.ai/mission
Proof — eval framework × UQ × weather-adjacent scientific ML × 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.
- Statistical methodology shape: distinguish real improvement from noise across basins, regimes, lead-time horizons (AMS multimodel streamflow; AMS 2021 floods & droughts) — same failure modes a weather LPM eval must surface.
- Reusable pipeline craft: Vertexium production agentic LLM + retrieval + multi-tenant agents; Lucent CTO monitoring; automated env-monitoring → validated reporting.
- Domain→metric translation: USGS/NOAA/NASA multi-source QA; hydroclimate / weather-adjacent research; publication-grade technical communication.
- Honest frame: central eval / UQ / weather LPM-adjacent lock — no Causal-internal results claimed; no WeatherMesh ownership; no invented pubs. Brand: the PhD who ships. SF onsite OK · Prefer take-home · Applied=0 until CEO GO.