May 2023 – May 2024 · 1 yr

• As an Advisory Board Member for the Data Science Club, I contribute my expertise and guidance to foster the growth and development of aspiring data scientists. With a passion for data-driven insights and innovation, I provide strategic direction and mentorship to the club, empowering members to excel in their data science journeys.

Jul 2022 – May 2023 · 10 mos

• During my tenure, we have promoted the field of data science by conducting a wide range of fun activities including hands-on workshops, 2 major events, unique projects made by the tech team, and real world problems being solved by the research department. The 2 major events include: Datahunt which was a Datathon with a twist of treasure hunt in it, and second was Model Mania were we wanted to show what AI and ML can do in a non-technical field so people can know its true power. Here we made creative models like Image-to-Art, AI Karaoke, AI joke generator and face mixer for anyone to try.

May 2021 – Jul 2022 · 1 yr 2 mos

• Worked on projects like sign language recognition using haar cascades and asteroid impact prediction on NASA's dataset to provide insights and predictions (Visualizations, Classification, Regression, Clustering along with deployment using FLASK).

Mar 2023 – Jun 2023 · 3 mos · Chennai, Tamil Nadu, India · Remote

• Planned and established the natural language understanding and generation
• pipelines. Built a multi‑level intent classifier model using the DIET classifier and
• performed hyper‑parameter optimisation to improve accuracy to 97%. In addition,
• fine‑tuned LLMs like Facebook OPT 2.7B to generate user‑specific solutions based
• on the user’s responses in real‑time along with tuning of inference parameters.

Nov 2022 – Feb 2023 · 3 mos · Chennai, Tamil Nadu, India · Remote

• The product that was being developed was a automated surveillance drone. I worked on the video analytics aspects of it which involved performing anomaly detection on live video and then also perform image captioning on anomaly frames so it is more explainable and user can clearly understand the situation.

Jun 2022 – Oct 2022 · 4 mos · Chennai, Tamil Nadu, India · Remote

• My main work involved around polygon matching problem. Given two polygon I had to find a mathematical approach to make those two polygons similar to each other using python. I got to work with QGIS software and .shp files and how we can incorporate math in computer science to solve real world problems.

May 2022 – Jul 2022 · 2 mos · Chennai, Tamil Nadu, India

• This was a research internship where initially I had to work on optimization of transformer models to make them production ready (Reducing Mode size and latency while maintaining accuracy). Later went on to work on a automated answer paper corrector where I used NLP to compare students and teachers answer and built a model to give the mark obtained as an output.