Data Scientist

Jio Platforms Ltd. (Reliance)

Aug 2023 - May 2024

Developed an agent-assisting, full-stack chatbot with an open-source Large Language Model, supporting Jio's dynamic knowledge base and integrating action automation within UX and infrastructure constraints.

  • Built chatbot leveraging open-source LLMs and Llama 2 for optimized performance and flexibility in the B2B use cases the company had.
  • Implemented automated actions and suggestions aligned with Jio UX requirements.
  • Ensured integration of cross-team requirements and hand-made guardrailing through various prompting strategies.

Research Consultant

Lexapar / JSW

Jul 2023 - Aug 2023

Focused on developing and enhancing Knowledge Graphs, Legal NLP, Named Entity Recognition (NER), and relationship extraction.

  • Consulted on custom knowledge graph development for legal data.
  • Worked with the team to build NER and relationship extraction methods for legal applications.

Applied Science Intern

Amazon, Alexa

Jan 2023 - Jun 2023

Conducted research on explainable AI and multilingual data generation for fairness in language processing.

  • Developed synthetic data augmentation (by generation) strategies to reduce bias and enhance fairness in Alexa concept clusters.
  • Generated explanations for underperforming clusters using methods ranging from TF-IDF to contrastive prompting.

Applied Science Intern

Amazon, Alexa

May 2022 - Jul 2022

Focused on data generation and augmentation for voice assistants, improving bias and fairness.

  • Designed data augmentation for improved fairness across multiple languages.
  • Conducted experiments to understand AI behaviour with explainable methods.

Intern Computational Linguist

Navana Tech

Oct 2021 - Apr 2022

Researched and implemented computational linguistics solutions, focusing on automated annotation and Indian phonology.

  • Scripted solutions to automate/speed up CL pipelines, Annotations etc.
  • Worked with speech team and training language consultants to capture dialectal variation and phonological richness in 9+ Indian languages.
  • Researched data to use, finding rules covering pattern-based errors, etc. and trained annotators on different tasks.

Intern Data Science & Linguistics

intelo.ai

Jun 2021 - Jan 2022

Developed NLP solutions for B2B CRM SaaS, specializing in keyword extraction, summarization, and sentence similarity.

  • Built tools for keyword extraction and summarization in CRM systems.
  • Implemented sentence similarity models to enhance CRM red flag finding.