work-experience

Machine Learning Engineer

Articulate.AI

• Responsibilities included
  • Spearheaded the development of advanced models for automatic patent description generation, utilizing cutting-edge long-context methods to handle complex and detailed patent data.
  • Integrated RoPE scaling techniques and optimized transformer architectures, enhancing the model’s capability to maintain context over longer sequences, crucial for accurate and coherent patent descriptions.
  • Fine-tuned state-of-the-art models such as Llama3.1, Mixtral 8x7b, Gemma 2, and Phi-3, achieving a significant increase in model performance metrics, leading to more precise and legally sound patent drafts.
  • Collaborated closely with cross-functional teams, including legal experts and patent examiners, to ensure that generated patent descriptions met industry standards and regulatory requirements.
  • Developed and deployed scalable machine learning pipelines for model training and inference, ensuring efficient handling of large-scale patent datasets and reducing processing time by 30%.
  • Contributed to research and development efforts, authoring internal whitepapers on the application of long-context models in intellectual property and patent documentation.

Associate Researcher

IIT Delhi

• Responsibilities included
  • Worked on projects in Computer Vision and Natural Language Processing for clinical detection of celiac disease and dialog systems.
  • Ideated and implemented Backend APIs for the Celiac Detection app, integrating features like Tissue Detection, Villi Tip Identification, Cell Counting, and Image Calibration using various computer vision algorithms.
  • Led a research project with three collaborators to develop a novel conversational dialog system, focusing on simultaneously identifying trigger points for recommendations and maintaining coherency and fluency in response generation.

Member Technical Staff

Oracle

• Responsibilities included
  • Analysing and automating transactional cases
  • Modelling modular framework for UI and REST Automation
  • Deploying Indsutrial Scale Systems
  • Creation of Validation Services

Machine Learning Intern

Grazziti

• Developed an end-to-end NLP engine for Topic-Modelling and Question-Answering
• Dataset included sentence, paragraph, and document level texts and created relevant and efficient search query extractions

Visual Intelligence Intern

Samsung R&D Institute, Bengaluru

• Part of the AI Computational Photography Team. Responsible for
  • Burst Image Denoising
  • Super Low Light Video Denoising