Jan 2024 – Jan 2026 · 2 yrs

• Involved in rapid development and prototyping of O1 and strawberry models.

Dec 2023 – Jul 2025 · 1 yr 7 mos · Gurugram, Haryana, India

• Final Selling Price (FSP) Prediction Model: Developing a machine learning model to estimate a car’s final selling price using fingerprint awareness, market science signals, and appointment-based condition assessments, having price accuracy of 91% at +-3.
• LLM -Based Inspection Summarization: Engineered an LLM-powered car inspection analysis system, extracting structured insights from car inspection reports and assigning condition scores (engine, body, tires, etc.), leading to a 4.7% boost in pricing accuracy.
• Default Sorting: Developed a new target variable for car sorting, achieving a 6.58% improvement in U2Bi.
• N1 Personalization: Engineered a personalized sorting, increasing N1 users' U2Bi by 31%.
• GA4 Migration: Led the migration of all models to Google Analytics 4, ensuring data fidelity

Oct 2023 – Dec 2023 · 2 mos · Bengaluru, Karnataka, India · Remote

• Developed a Named Entity Recognition (NER) system to extract meaningful information from documents like invoices and financial statements, leading to a business of 230Cr.
• Created and trained cutting-edge NLP models using transfer learning from existing transformer architectures and fine-tune them for specific use cases, such as sentiment analysis, topic modeling, and information extraction

May 2023 – Sep 2023 · 4 mos · Remote

• Created a library for training state-of-the-art NLP models which uses transfer learning from the existing transformer architectures and fine tune on a specific use case.
• Each of the transformers adds neural network layer(s) at the end, which is then trained on the supplied data.

May 2021 – May 2023 · 2 yrs · Remote

• Developed an isolated service that facilitates the comparison of Logos using the VGG16 model. The module reduces the overall False Positive rate by 4% while incurring only 2.85% of False Negatives.
• Improvised the current Function and Division mapping algorithm by improving discovery and including 1-to-many mapping, increasing the number of functions from 12 to 15 and divisions from 57 to 198.
• Created a POC BERT model which overcomes the need for rules-based mappings to extract tech from company/account pages; as a part of A/B test configuration (A=Rules Based / B=BERT based model) to gauge confidence and look to replace it when enough confidence is achieved.

Feb 2023 – May 2023 · 3 mos · Department of Science and Technology, New Delhi · On-site

Dec 2021 – Feb 2023 · 1 yr 2 mos · Department of Science and Technology, New Delhi · On-site

• I have been involved in several tasks pertaining to data mining, applying scientific methods to analyze it objectively and providing evidence-based insights for DST’s ongoing schemes. Moreover, I have also extracted data generated from 28 programs under four umbrella schemes which were scattered sources, crystallizing the massive information volume into answers for complex questions that provided country and program-wide insight.
• Under the guidance of DST’s Chief Data Officer, Dr. Akhilesh Mishra, and the Project Monitoring Unit, I have also engaged in creating an interface for DST to capture project-based data from various divisions in a standardized format. I collaborated with DST’s NMICPS division and helped them to streamline and visualize their sanction-based data to identify the fund flow.
• I helped create compatible data for designing interactive Tableau dashboards for the DST, providing insights on its ongoing schemes.

May 2022 – Apr 2023 · 11 mos · Jaipur, Rajasthan, India

• • Elected representative for the student body, acting as the POC between the students, faculty and the management.

Jul 2021 – Jul 2022 · 1 yr

• Responsible for execution of all the society’s undertakings.
• Managed 4 workshops, 2 Inter-college competitions, and 2 awareness drives.

Aug 2019 – Jul 2021 · 1 yr 11 mos

Aug 2018 – Aug 2019 · 1 yr

May 2021 – Jul 2021 · 2 mos

Dec 2020 – Dec 2022 · 2 yrs · New Delhi, Delhi, India · On-site

Apr 2020 – Aug 2020 · 4 mos · Seattle, Washington, United States

• Worked on the live CC-TV footage and integrated multiple Screen cc-footage video data using Computer Vision techniques.
• Worked on object recognition and tracking, image registration, image calibration, and correction.

Jan 2020 – Mar 2020 · 2 mos · India

• Worked on experimentation and rapid prototyping.
• Helping the company to take my R&D through to production.

Apr 2019 – Jun 2019 · 2 mos · Greater Delhi Area