Software Development Engineer
May 2026 – PresentWavexcel Technologies · Client: Siemens Energy
>90%faster turbine design — 3–4 hours down to 10–15 minutes
36real Siemens Energy projects validated against expert designs
>50%faster AutoCAD drawings — 36 drawings in about 20 minutes
- Engineered an AI-based steam-turbine design optimizer tuning 15 variables under 16 constraints, cutting design time from 3–4 hours to 10–15 minutes (>90% faster) across a 5.4 × 1016 search space.
- Implemented constrained Bayesian optimization (Gaussian-process surrogate + feasibility model) with a dependency-matrix prior, integrated end-to-end with Siemens’ solver toolchain (Turba / WelleSST / RSmin) via automated file I/O and parsing.
- Validated across 36 real Siemens Energy projects, matching expert designs and exceeding senior engineers’ efficiency on multiple cases, aided by a self-learning cross-project warm-start library.
- Automated AutoCAD drawing generation in the turbine piping module, driving each drawing from the prior stage’s inputs and outputs — from 60s to 28s per drawing (>50% faster), 36 drawings in about 20 minutes.
- Automated 179 datasheet cell computations from 31 input drivers via a formula-driven cascade engine, achieving a 100% pass rate on 17 tests.
C#.NET MAUIPythonBayesian OptimizationGaussian ProcessesAutoCAD Automation
Microservices
JWT