SHOUNAK LAHIRY

Building digital solutions to improve the accuracy of real time demand to supply matching algorithm thus improving customer and partner experience on Ola platform in India

Key Projects:

• Facilitated unified network for ride hailing, food delivery and logistics businesses with a common electric 2W supply optimizing for their earnings, utilization, charging time and last mile delivery metrics
• Implemented intelligent allocation system that takes into account the battery charge status of electric 2W, resulting in a 15% improvement in vehicle uptime thereby a 12% increase in revenue within category
• Built a dynamic fetch system in allocation algorithm using isochrones for different h3 indexes, this was aimed at retrieving most relevant drivers for each booking request, resulted in 10% improvement in driver cancellation
• Improved allocation system by transfer legacy algorithm that considered linear cost function with limited feature to a more dynamic and real time multi-feature of demand and supply based model thereby increasing completed rides by 4%
• Led initiatives to optimize first-mile distance for driver partners in the 2W category, successfully reducing first-mile distance by 1 km per ride in top 7 cities in India
• Addressed customer dropouts due to stockouts & cancellations by introducing upsell to higher categories .Achieved a 2.5Cr monthly GMV increase and raised ride completion rate by 3pp.
• Optimized allocation algorithm for Parcel delivery leading to improvement in net delivery fulfillment by 3% DoD for bangalore
• Currently working on customer attributes (CLM) in network optimization aiming to improve customer driver matching and reduce allocation time while maintaining partners’ preferences and utilization

Product Analytics backed initiatives:

• Led initiatives to improve supply density at micromarket level for cabs category by analyzing key demand and supply trends in terms of distance of supply from demand hotspots
• Reduced driver cancellations by 15% during ongoing trip by improving allocation strategy that assigns driver who are intrip when a booking request is generated
• Introduced Freeze timer feature in 3W, intelligent allocation system that selects drivers who have maximum willingness to complete a booking request. This led to incremental 1.2cr in monthly GMV for 3W PAN India

Worked on developing precise strategies in pricing domain to increase customer conversion on Ola platform using data science and applied research

Key Projects:

• Transformed legacy surge pricing algorithm by building a model to predict supply acceptance using near real time driver performance and demographic based features as well as predict real time customer conversion using real time customer behavior
• Facilitated extension of surge pricing algorithm specifically for 2W category using machine learning models specific taking demand and supply context for ride haling, food delivery and parcel delivery.
• Developed customer base pricing strategy for electric 2W launch in India by working backwards from unit economics of driver and use case that customer would choose electric 2w leading to jump in new customer acquisition by 5% post launch
• Led initiative to reduce allotment time in 4w category by introducing retry level dynamic driver earning change by optimizing allocation algorithm at different stages of booking cycle
• Devised drop destination based pricing strategy for cabs which takes into account demand and supply dynamics of destination geohash, this led to improvement in cabs GMV by 10% week on week for top 7 markets in India
• Increased driver acceptance for high first mile distance rides by around 15% for cab category by introducing feature call CAB (contextual allocation booking) that dynamically adjusts driver incentive for these scenarios during booking conversion cycle itself

Working with the Business Finance function to monitor financial health and drive cost and margin improvement programs

Key Projects:

• Single-handedly managed business planning and reporting, root cause analysis and insights for weekly business review meetings using smart and interactive dashboards, big data engineering tools
• Reduced driver acquisition cost by 3% by restructuring vendor pay-out schemes
• Lowered customer acquisition costs by 5% by developing customer LTV model

Led teams (sized 15-30) to build analytical roadmaps and implement data driven solutions for retail, tech and FMCG clients

• Identified future consumers for a product by using decision tree and rule-based heuristics to classify consumers into “High/Medium/Low” usage potential for a $110B technology giant to build social media intelligence and an analytics framework for their new B2B product
• Delivered analytical solutions in sales performance and campaign effectiveness to a $35 B chocolate and pet care manufacturer

• Developed an automated tool to identify anomalies in product trends with insights on promotions, new launches
• Generated company business of ~$0.5 M by identifying drivers of negative growth in sales for a customizable footwear division