Jan 2022 – Dec 2024 · 2 yrs 11 mos · Bengaluru, Karnataka, India

Jan 2021 – Jan 2022 · 1 yr

Apr 2020 – Jan 2021 · 9 mos · Gurugram, Haryana, India

• Part of the Data Science team towards development and deployment of various machine learning models
• Website Recommendation
• For personalised user recommendation, similar listings, popular listings & identifying potential buyers for a listing (Showroom
• Visit Predictor) to be used by the Sales Team resulting in 20% decrease in filter usage and 25% increase in lead conversion

Feb 2019 – Jan 2021 · 1 yr 11 mos · Mumbai, Maharashtra, India

• Lead the Data Science team towards development and deployment of various machine learning models
• Car Pricing
• Achieved an error rate of < 5% for 80% of the major car models in price prediction using a very small training dataset - Reduced the human dependency for calculating car prices which streamlined the process and removed human biases
• Car Dynamic Pricing
• Using TruePrice as a benchmark, changed the prices dynamically to adjust as per the current demand on the website
• Increased monthly Post Tax Margin from 3% to 7% across all cities without increasing Inventory Holdup
• Marketplace Car Cannibalisation
• Removed 10% marketplace cars from the platform that resulted in Inventory cars (20% of total cars) receiving 65% total leads
• on the website, as opposed to 30% earlier without increasing the average lead cost
• Procurement Score and New Car Store Assortment
• Decreased average inventory holdup at the stores from 25 to 21 days and decreased bad car procurement by half

Feb 2017 – Feb 2019 · 2 yrs · Mumbai, Maharashtra, India

• Worked in the Model Risk Management team to independently validate various risk models across market risk, credit risk and regulatory capital calculation.
• Investigate all aspects of the model covering their assumptions regarding the behaviour of different asset classes, performance over different time scales, treatment of different products and risk types and their aggregation and use in the risk capital calculations.
• CCAR | Home Price Index (HPI) Scenario Expansion Model
• Used to predict Home Price Index estimation for all US states
• Performed OLS assumption testing, outlier estimation, back-testing, benchmarking and sensitivity analysis for the model
• Predicted the model output for various stress scenarios in the future as per FRB guidelines
• Advanced Internal Rating Based (AIRB) | Global Hedge Funds PD Model
• Used to calculate credit rating for all CS Hedge Funds counter-parties using Expert Rank Ordering (ERO)
• Independently reconstructed the model using Single and Multi-Factor Analysis
• Developed algorithms in R to assess potential model risk
• Performed statistical tests on the performance and assumptions
• Economic Risk Capital (ERC) | Collateralized Loan Obligation (CLO) Warehouse
• Used for calculating the risk in funding the purchase of loans
• Challenged the methodology for shock calculation with bootstrapping
• Performed sensitivity analysis at various shocks

Dec 2015 – Apr 2016 · 4 mos · Kanpur, Uttar Pradesh, India

• For the course "Digital Electronics" offered by the Department of Electrical Engineering, IIT Kanpur

Jul 2015 – Nov 2015 · 4 mos · Kanpur, Uttar Pradesh, India

• For the course "Electrical Engineering Lab" offered by the Department of Electrical Engineering, IIT Kanpur

Aug 2015 – Feb 2017 · 1 yr 6 mos · Kanpur, Uttar Pradesh, India

• The research was supported by the National Centre for Flexible Electronic. Our work has applications towards increasing efficiency of flexible solar cells made using Organic molecules.
• Particularly, we worked on testing the effects of Annealing on Ternary Planar-Bulk Hetero-Junction Organic Solar Cells.
• Increased the efficiency from 1.1% to 3.93% by optimising annealing temperature and streamlining the fabrication process.

May 2014 – Jul 2014 · 2 mos · Hyderabad, Telangana, India

• Modelled & developed a test-bench to test the coverage, accuracy & efficiency of algorithmic memory access.
• Analysed various bank and address conflicts in Verilog, making it scalable for various port capabilities.
• Designed System Verilog cover-groups for testing efficiency of the developed Verilog test-bench.
• Verified memory structures with the test-bench, which helped approve memory designs.

May 2013 – Jun 2013 · 1 mo · Kanpur, Uttar Pradesh, India

• Built a Hall sensor circuit for Hindustan Aeronautics Ltd. to measure temperature and RPM of military aircraft wheels.
• Conceptualised a teeth assembly to be mounted on the wheels, which would be sensed by the hall sensor and fed to a Butter-worth Filter to convert digital output of sensor to a sinusoid wave to its frequency.
• Mounted the assembly on a Maxon DC motor and compared its performance with its internal encoder.