Experience

Jan 2022 - to present

Senior Data Scientist, Simulation & Optimization

Sahil is a Senior Data Scientist at Walmart’s Data Science & Strategy Org. He leads the development of Simulation-based Digital Twin and Advance Optimization Algorithm work for Walmart fulfillment centers and Ecomm assets. And serves as a simulation consultant to internal Walmart teams. Sahil has developed and deployed multiple detailed simulation models and machine learning-based optimization algorithms for existing and upcoming Walmart Fulfilment centers. These models have transformed how Walmart optimizes their asset planning, technology introduction, business strategy, customer experience, operations, and resource planning.

Strategy and Consulting

Leading research team on simulation-optimization projects for Walmart’s Next Generation Automated Delivery Fulfillment Centers, guiding business decisions on technology diffusion to enhance associate and customer experiences.
Defined roadmap and identified opportunities for Simulation and ML optimization projects for Walmart PDCs, guiding End-to-End solution implementation to drive Middle Mile transformation engineering decisions.
Leading efforts to architect an Unified Simulation Platform for on-boarding Network-Wide simulation models.
Mentoring teams to productionalize Simulation tools infrastructure across Strategy and Operations teams, replacing Excel models with detailed simulation testing sandbox for process optimization.
Actively involved in leadership decisions on next-generation Walmart projects, impacting supply chain optimization for ~4,000 stores.

Technical Contributions

Architected 5+ digital twin models with Machine Learning based self-optimizing capabilities for end-to-end process optimization.
Deployed models are actively used to optimize picking, bot routing, last-mile delivery, resource scheduling, layout planning, and asset throughput.
Modeled FlexSim & SimPy based Value Stream simulations, for A/B testing Automation design concepts and process improvement plans. Facilitated a 90% faster validation turnaround time and improved accuracy in KPIs by 33%
Architected 7+ new FlexSim objects using custom code to model Path Planner toolbox, OEE tracker, Loading docks assignment, custom Workstations, ERP system connections, enhancing feature availability of software.
Developed a 'Layout Evaluator Simulation' using FlexSim and NVIDIA Omniverse to reduce flow congestion. Eliminated manual mock-ups, accelerating the layout selection process and saving 2 months & 300+ labor hours.
Patented an innovative inventory placement strategy and SKU selection algorithm. Developed the concept of ‘Holistic business metric rank’ for SKU selection that increased asset throughput, and pick accuracy .
Innovated a SimPy based ML routing simulation-optimization frameowrk based on vicinity clustering which increased pick rate by 17\% resulting in cumulative savings of 160K USD, decreased travel distance (~1 mile per bot), and faster order turnaround time.
Introduced data-driven decision-making capabilities by democratizing simulation tool infrastructure across design, strategy, operations, and product teams, replacing traditional Excel-based models with detailed simulation sandbox access.

Dec 2019 - Dec 2021

Research Assistant

Sahil graduated from Northeastern University with a Master's in Industrial Engineering with a focus on Machine Learning for industrial optimization. For 2 years Sahil worked in the Northeastern University Simulation Lab as a Research Assistant. During which he has delivered meaningful research projects, published papers, developed new curriculum, and collaborated with key organizations like Simio and Mathworks. His research was focused on developing and testing the applicability of Reinforcement Learning algorithms for industrial applications such as Supply chain, warehousing, Robotics, Queueing networks, and UAVs.

Collaborated with Simio to model an ERP-based, real-time Discrete Event Simulation multi-objective optimization of a Digital Twin model for a Beverage Manufacturing Plant.
- Research provides insights on a framework to link commercial simulation software with Python-based optimization algorithms.
- Implemented API integrating Simio, Excel, and Python to apply NSGA-II multi-objective model for optimizing production scheduling process.
- The sim-opt framework resulted in a 25% increase in throughput with an 11% reduction in total operating cost when dispatching rules and operator schedules are updated according to the suggested simulation results.
Conducted an exhaustive review of research trends in using Reinforcement learning for industrial applications.
- Authored 6 presentations and published 2 review papers explaining implementation feasibility, technology diffusion, and benchmarking of simulation optimization using Reinforcement learning for manufacturing, supply chain, and operations research applications.
Developed Reinforcement learning-based optimization model for job shop scheduling problem. The paper published describes the methodology and steps for defining a reward function for JSSP.
Innovated simulated annealing-based optimization algorithm for UAV routing with charging station application for drove delivery applications. The paper benchmarks algorithm performance and benefits over known heuristic methods.
Developed a university-level course for Mathworks to teach simulation and Reinforcement Learning using MATLAB for engineering applications. The course currently has 750+ downloads and active usage in multiple universities.

Jul 2020 - Dec 2020

Manufacturing Engineer

Managed 4 projects with cross-functional teams (operations, maintenance, RnD) to achieve daily production targets, using QDIP scorecards and Pareto analysis.
Improved the OEE of 9 work-stations by 34% implementing Failure Analysis, Root Cause Investigations, Design of Experiments (DOE), and A3 plan for 18 robotic work stations across 2 lines.
Validated a $40M FDA standard clean-room manufacturing line by conducting series of qualifications (IQ, OQ, PQ) with a team consisting of automation vendor, operations, supplier engineering, and quality.
Documented 11 work instructions (MAXIMO), maintained 4 manufacturing aids (Arena Solutions), executed 7 Process Change Orders to reduce scrap rate on 2 assembly lines by an average of 20%.

Jan 2018 - Jul 2019

Research Engineer

Co-established and incubated $1M Manufacturing & Testing Facility for Industrial Heat Pumps and Chemical Separation Membranes.
Led product research and operations to streamline manufacturing process across 3 facilities around the globe.
Spearheaded product identification and positioning process to establish, validate and, identify product placement in 16 industrial sectors across India.
Successfully transitioned the start-up to an SME company, which was later acquired into ChemDist group of companies .