Sep 2023 – Oct 2024 · 1 yr 1 mo

• Deployed and Achieved 54% automation of customer support tickets end to end with RAG based LLM autoresponse saving 130 hours in total everyday for 70 customer agents.
• Improved loan collections by 1.2 crores per month using SMS based analytics and intelligent collection efforts based on repayment trends.
• Tech Stack used: Python, FastAPI, OpenAI API,

Apr 2023 – Aug 2023 · 4 mos

• Successful Built a personality based classification model prototype from scratch for internal use and successfuly proved the utility in loan credit and collections behaviour for future applications

Sep 2021 – Aug 2023 · 1 yr 11 mos · Bengaluru, Karnataka, India

• Using Deep learning techniques to build a texture reconstruction model to derive original presented stimulii from the associated neural signals shown to macaque monkeys for brain machine interface applications.
• Lead Data Scientist for two major dementia studies (TLSA and SANSCOG).
• Published and contributed towards multiple research manuscripts by handling statistics and deriving insights.
• Handled all data based operation decisions for study design and execution

Aug 2020 – Aug 2021 · 1 yr

• Optimized and deployed supply chain forecasting methods and improved sales by 40 Lacs./month
• Designed and deployed from scratch with no precedence supplier selection algorithm for the whole company and increased fulfillment by 4%.

May 2020 – Jul 2020 · 2 mos

• Developed an action recommendation system for increasing generic medicines sales for the company by integrating criterias such as customer behaviour, market demand etc.
• Developed a tableau dashboard accompanying the above algorithm for easy usage and interface for stakeholders

Aug 2019 – Apr 2020 · 8 mos · Rome

• Classification of elementary particle images for neutrino experiments.
• Major challenges included dealing sparse data and bypassing manual bounding box specification in State of the Art

May 2019 – Jul 2019 · 2 mos · Mumbai Area, India

• 1. Optimisation of Calling Pipeline to generate more loan conversions using Machine learning
• 2. Product categorisation for targeted marketing of customers using Machine Learning.

May 2018 – Jul 2018 · 2 mos · Pilani

• Worked on improving state of the art accuracy for 4x Super Resolution of an Image using Deep Learning.
• Experimented with different combinations of state of the art models for Super Resolution such as SRGAN, SRCNN, EDSR, DBPN etc. to improve the metrics.
• Achieved RMSE 11.46 and Perpetual Index 5.89
• Participated with a team of four in Perpetual Image Restoration Challenge held in conjunction with ECCV 2018 conference and secured 13th rank on global platform.
• Technologies used: Python with Tensorflow and Pytorch.

Dec 2016 – Jan 2017 · 1 mo · Bangalore

• 1. Worked at SWAN program of government of India for promoting radio astronomy throughout the country.
• 2. Went to Gauribidanur Radio Astronomy observatory to learn SWAN operation.