Sep 2022 – Nov 2024 · 2 yrs 2 mos · Gurugram, Haryana, India · On-site

• MeraPashu360 is a game-changing platform that is dedicated to enhancing the lives of millions of dairy farmers across India. Our omnichannel marketplace offers farmers access to high-quality cattle, nutrition (cattle feed), and veterinary services, all of which were previously difficult to come by. We are disrupting the unorganized cattle trade market by ensuring quality assurance of cattle being traded on the MeraPashu360 platform.
• We are revolutionizing the lives of dairy farmers by leveraging the power of data science to create cutting-edge solutions. We have developed ML-based models for cattle-trade pricing, computer vision-based tools for cattle health determination, and Business Intelligence/Competitive Intelligence/Analytical dashboards for understanding customer preferences and the market landscape. Additionally, we have implemented optimization-based solutions for warehouse location planning, inventory management, and logistics planning, which have significantly improved the efficiency of our operations. To support these we have established data warehouse (using Big Query) and tailored efficient data pipelines to migrate all online/offline data automatically to our data warehouse.

Aug 2021 – Sep 2022 · 1 yr 1 mo · Bengaluru, Karnataka, India · On-site

• As a data science leader, I spearheaded the development of cutting-edge, data-driven solutions for Ola Electric and Ola Cars. These solutions have enabled the companies to optimize their operations, improve customer satisfaction, and drive growth.
• My contributions includes:
• A machine learning-based model for predicting the range of electric vehicles
• Utilization of EV telematics data on the cloud to predict and analyze faults and customer driving behavior
• Implementation of voice assistants on electric vehicles using ASR, NLP, and TTS technologies
• Deployment of a RASA-based chatbot for user support
• Utilization of NLP for social media management to understand and engage with customers
• Automation of EV factory processes utilizing inventory management, supply chain, and optimization techniques
• Party Mode functionality on Ola Scooters
• An ML-based pricing model for used cars
• A computer vision-based solution to detect damages in cars
• Customer segmentation for targeted marketing campaigns
• A recommender system for listing cars based on user preferences

Aug 2020 – Aug 2021 · 1 yr · Bengaluru, Karnataka, India · On-site

• As the leader of a talented team of operations research scientists, data scientists, and software developers, I guided and mentored them to design and develop optimization-based software solutions. Throughout the end-to-end software development cycle, I ensured that all stakeholders were kept informed and the project was delivered on time and within budget.
• One of our major achievements was implementing the RailMax solution for Indian Railways, which helped them generate optimized train paths for running freight trains efficiently. This planning and operation tool has enabled users to monitor and visualize their current operations, ultimately leading to more efficient freight train schedules.
• Another notable project we worked on was optimizing train schedules for BHP Iron Ore in Australia. As they own a private rail network to move iron ore from mines to port, we designed a solution that maximized the throughput of their constraint rail network. The BHP planners use this solution on a daily basis to prepare optimized train departure schedules, leading to significant improvements in efficiency and profitability.
• Throughout these projects, I leveraged my expertise in operations research, software development, and project management to deliver results that exceeded client expectations. My team and I worked collaboratively, embracing challenges and finding innovative solutions to complex problems.

Jan 2018 – Aug 2020 · 2 yrs 7 mos · Bengaluru, Karnataka, India

• As a data science and optimization leader, I spearheaded multiple projects with significant contributions such as:
• Ancillary Dynamic Pricing Engine, Client: Etihad Airways:
• We designed and built a machine learning-based solution that helped Etihad Airways determine optimal prices for ancillary services like preferred seat selection and excess baggage. We utilized PySpark to develop the solution on the AWS cloud. My responsibilities included defining the project requirements, leading the development of the solution, and coordinating the deployment of the solution into the production environment.
• Dynamic Availability, Client: Multiple Airlines:
• We developed a machine learning-based model to predict the probability of converting a flight search to booking based on various attributes like passenger, flight, and price, etc. We used this model to build the Dynamic Availability solution, which helped airlines monitor market context in terms of prices offered by their competitor airlines and dynamically adjust their offered prices. I was responsible for preparing ROI studies for potential clients, performing value estimation studies for understanding the impact of changing the source of fares, changing the way robotic shopping is done, and frequency at which fares are changed.
• Revenue Optimizer, Client: Multiple Airlines:
• We developed Revenue Optimizer, a revenue management tool that helped airlines price their seats optimally by analyzing historical purchase behavior and demand patterns. As a team lead, I managed the implementation of the Revenue Optimizer at multiple airlines, designed several enhancements to the existing solution, and built prototypes in C++/R to measure the value addition by those enhancements. I also performed simulation-based detailed ROI studies for potential clients.

Jul 2015 – Dec 2017 · 2 yrs 5 mos · Bengaluru, Karnataka, India

• Built / delivered cutting-edge solutions for multiple airlines. Here are some examples of the projects I worked on:
• Flight Disruption Management System, Client: Multiple Airlines
• I led the development of an innovative solution that automates the re-accommodation of passengers in the event of a flight disruption or cancellation. Our system was calibrated to meet the specific needs of each airline we worked with, resulting in high adoption rates and increased customer satisfaction.
• Optimization-Based Cargo Loading System:
• I designed and implemented an optimization-based solution to automate cargo loading operations for airlines. Using Mixed Integer Programming techniques, we were able to determine the optimal loading scheme for the aircraft's belly cargo. We fine-tuned the parameters in the model to meet each client's unique requirements.
• Cargo Network Merger Study, Client: Etihad Partner Airlines
• Etihad engaged our team to evaluate the feasibility of merging the cargo networks of their partner airlines. We built a sophisticated tool using Discrete Event Simulation techniques in C++, which allowed us to simulate cargo operations for each airline. We performed various studies, validated and benchmarked them, and presented our recommendations to the Etihad Executive team. Most of our recommendations were adopted and successfully implemented.
• My work also extends to Machine Learning and Artificial Intelligence techniques, and I have successfully implemented/enhanced solutions such as dynamic pricing engines for airlines and optimization-based revenue management tools.

Jan 2012 – Apr 2015 · 3 yrs 3 mos · Gainesville, Florida, United States

• As the lead for Rail & Mining Industry solutions, I spearheaded the development of a suite of tools that revolutionized the way our clients conducted their operations. One such solution was the Network Simulation System (NSS), which helped a major iron ore mining company to optimize their supply chain by simulating their operations and identifying bottlenecks. I worked closely with the company executives to understand their requirements and led a team of over 10 members to build an insightful decision support system that utilized the latest DES techniques. Our tool provided rich animation screens with detailed KPIs and was instrumental in finalizing the plans for capital expansion.
• Another notable solution that we developed was the Real-time Rail Scheduling Tool (RST), which enabled a rail company to create optimized train schedules and efficiently utilize their network. As the lead, I worked closely with the client to formulate, clarify, and expand business and technical requirements. With a team of over 20 members spread across 5 time zones, we skillfully designed and deployed the tool in under a year, providing real-time visibility of current operations while creating an optimal departure plan. Our innovative solution not only improved the client's efficiency but also helped them stay ahead of the competition.

Oct 2010 – Mar 2012 · 1 yr 5 mos · Gainesville, Florida, United States

• As the leader of the team, I was responsible for designing and developing HaulPlan, a cutting-edge software solution for Less Than Truckload trucking companies. HaulPlan helps trucking companies create optimized operational plans that include Loading Plans, Departure Schedules, Driver Routes, and empty trailer re-positioning.
• Some of my key contributions to this project include:
• Designing the underlying mathematical model to optimize the operational plan for LTL companies, taking into account factors such as capacity constraints, delivery times, and distance.
• Creating a Graphical User Interface (GUI) that is user-friendly and intuitive, enabling trucking companies to easily visualize and manage their operational plans.
• Conducting Proof of Concept (PoC) studies and Return on Investment (ROI) analyses for potential clients, to showcase the effectiveness of the solution and demonstrate its value proposition.
• Delivering the solution to the client, working closely with them to ensure a smooth deployment process and providing ongoing support to ensure optimal performance.
• HaulPlan has been well received by our clients, enabling them to optimize their operational plans, improve delivery times, reduce costs, and increase overall efficiency.

Jun 2008 – Sep 2010 · 2 yrs 3 mos · Gainesville, Florida, United States

• My expertise in OOPs and data structures allowed me to create robust and efficient solutions that met the unique needs of each client and helped them achieve their business objectives.
• Developed and implemented advanced algorithms and data structures using object-oriented programming principles (primarily in C#) to support a range of Decision Support Systems.
• Led the design and development of a Crew Scheduling Optimizer for Class I railroads, a powerful software tool that creates optimized crew assignments for rail schedules while adhering to all relevant regulations. This involved designing and implementing an OR mathematical model and developing the necessary database architecture (SQL-based) to support the solution.
• Built a Service Validation and Flow Writing Tool for Parcel Delivery Company, an innovative solution that enables large-scale parcel delivery companies to validate their service commitments and assess the effectiveness of their plans. Leveraging graph theory and other advanced techniques, I designed and developed an efficient algorithm to meet the project's business objectives using C#.
• Served as a key contributor to the "One Plan" consulting project for a major Class I railroad in the US, playing a critical role in all stages of the project, from data gathering and analysis to implementing extensions for Optym's Train Scheduling Optimizer tool to meet the project's needs. Ultimately, I helped present the results of the optimized long-term plan (including train schedules, blocking plan, and crew schedules) to the client.

May 2007 – Aug 2007 · 3 mos · Delhi, India

• As a part of the technical consulting team for Bharti Airtel, I contributed significantly towards the development of a heuristic model for performing social network analysis on a massive communication network. My primary objective was to identify potential clients and possible churn outs. The volume of data was huge, encompassing every call record over the past few years, and required the use of various big data processing techniques.
• To achieve the desired outcome, I employed advanced data structures and custom algorithms implemented in Java. The solution involved designing and implementing a robust framework for processing the data in real-time and extracting meaningful insights.
• Throughout the project, I worked closely with the Bharti Airtel team, leveraging my technical expertise to provide guidance and support. The project was a great success, providing valuable insights to the client and contributing significantly to their bottom line.

May 2004 – Aug 2004 · 3 mos · India · On-site

• As a key member of the engineering team, I designed and developed a simulation software to evaluate the potential benefits of a new assembly line design for trucks. This software was built using advanced simulation techniques and algorithms to create a virtual model of the proposed assembly line.
• Working closely with the manufacturing team, I gathered data and analyzed the current assembly line process, identifying areas for improvement. Using this data, I developed a new assembly line design that would improve efficiency, reduce costs, and increase throughput.
• The simulation software allowed us to test this new design in a virtual environment, modeling the entire assembly process and simulating various scenarios to identify potential issues and optimize the process. We used advanced visualization tools to provide real-time feedback, allowing us to adjust the design and evaluate its performance in real-time.
• After extensive testing and optimization, we were able to demonstrate significant improvements in efficiency and throughput, validating the new assembly line design and providing a clear path for its implementation in the manufacturing process. This project was a key factor in reducing costs and increasing productivity for the company, contributing to its ongoing success in the market.