Anurag Soni

I’m an Operations Research & Optimization student who enjoys turning complex, real-world constraints into measurable improvements. My work sits at the intersection of optimization, stochastic modeling, and simulation, with a strong pull toward learning systems that adapt optimization behavior over time.

In projects, I’ve implemented and analyzed Rao family parameter-free optimization algorithms (BMR, BWR, BWMR), benchmarking them against classical metaheuristics on standard benchmark functions. I’m also exploring how machine learning can support adaptive optimization rather than relying on fixed heuristics.

I built a stochastic manufacturing system simulation in SimPy using MTBF/MTTR with buffering, modeling Poisson arrivals and normally distributed service times. By analyzing utilization and failure behavior to pinpoint bottlenecks, I achieved a 61.8% productivity improvement through system redesign—an outcome driven by data-backed operational reasoning.

I’ve also tackled supply chain optimization via linear programming, modeling multi-plant, multi-region systems and solving with Python (PuLP) while validating results with Excel Solver. Beyond operations work, I developed an AI-based video processing and OCR pipeline using CNNs, LSTM-based reinforcement learning for OCR correction, and a Flask backend with database integration, reinforcing how I approach engineering problems end-to-end. I bring leadership through organizing at scale—leading 30+ members and securing 75K+ for initiatives—while staying focused on rigorous analysis and practical impact.