You can download my CV here. (Last updated 30th Jul. 2024)

Education 🎓

- University of Glasgow, Glasgow, UK
  - Master’s Thesis Research Student, Grade: Excellent, [Dec. 2023 - Mar. 2024]
  - Supervisor: Dr. Chaitanya Kaul, Research Associate, School of Computing Science.
  - Thesis: Advancing Medical Image Segmentation Through Multi-Task and Multi-Scale Contrastive Knowledge Distillation

- Birla Institute of Technology and Science, Pilani, India
  - Master of Technology in Data Science and Engineering; GPA: 7.75/10, [Apr. 2022 - Jun. 2024]
  - Courses: Advanced Mathematics, Data Mining, Statistical Modelling, Machine Learning, Deep Learning, Computer Vision, NLP.

- Siliguri Institute of Technology, Maulana Abul Kalam Azad University of Technology, India
  - Bachelor of Technology in Computer Science and Engineering; GPA: 8.84/10, [Aug. 2014 - Jun. 2018]
  - Departmental Rank - 1
  - Supervisor: Prof. Anupam Mukherjee, Head of the Department, Dept. of Computer Science & Eng.
  - Thesis: Autonomous Self-Driving Bot: Real-Time Obstacle Detection and Decision Making
  - Courses: AI, Data Mining and Data Warehousing, Discrete Mathematics, Probability, Data Structures, Design and Analysis of Algorithms.

Skills 🧠

- Languages: C, C++, Python, PySpark, SQL
- Frameworks: PyTorch, TensorFlow, Keras, Scikit Learn, LangChain, NLTK, SpaCy, OpenCV, Django, Flask
- Tech Skills: Deep Learning, Computer Vision, NLP, Generative AI, Vector Database, Time Series Forecasting, Convex Optimization
- Platforms: Linux, Windows
- Soft Skills: Leadership, Team Management, Public Speaking, Mentorship

Experience 👨🏻‍💻

- Optiks Innovations Pvt. Ltd. (P360), Remote
Principal Engineer - Artificial Intelligence, [Mar. 2024 - Present]
  - Leading a team of data scientists and machine learning engineers to tackle advanced R&D challenges in pharmaceutical commercial sales operations.
  - Architected and deployed EdenHelp, an industry-first Conversational AI platform to optimize medical inquiries for pharmaceutical drugs and medical devices.
  - Developed the understanding and generative AI components, ensuring hallucination-free responses using advanced NLP, RAG, and knowledge graphs.
  - As part of my functional roles, I do project planning, sprint planning and management, help in building RFPs for business development from an Advanced Analytics/AIML perspective.

Technical Lead – AI, [Mar. 2021 - Mar. 2024]
- Responsible for building ML models for BirdzAI, an AI powered data analytics for precision pharma sales & marketing.
- Designed and developed a technique blending supervised and unsupervised model to predict physician prescription behavior and loyalty for a leading gastrointestinal product in USA.
- Designed and developed a prescription demand forecasting model using Gaussian Processes, overcoming challenges in designing the optimal kernel through rigorous hyperparameter tuning, cross-validation, and simulation, resulting in a 15% improvement in current forecast accuracy and reaching an overall accuracy of 93%.
- Designed and developed a multi-class text classifier using RoBERTa and SVM for intent classification, reaching an accuracy of 94% with 46 classes.
- Sentiment Analysis on conversational data using Transformers based Models.
- Created a collaborative-filtering based recommendation model for omnichannel engagement of physicians and their staffs.

- P360, Piscataway, New Jersey (Remote)
Data Scientist, [Mar. 2019 — Mar. 2021]
- Worked on Call Planning Optimization using Ensembled Classification Algorithm and Propensity Scoring.
- Worked on creating an algorithm for Detecting Acquired Blepharoptosis in Human Eye using Facial Landmark tracking and Deep Neural Networks.
- Built Marketing Mix Model for better Omnichannel and Customer Engagement – Panel Data Regression for Time Series Forecasting using linear models.
- Built Physician Segmentation Model using unsupervised techniques to do Customer Profiling in real-time.

- Prescriber360 Solutions, Piscataway, New Jersey (Remote)
Associate Data Scientist, [Mar. 2018 — Mar. 2019]
- Information retrieval using Optical Character Recognition (OCR) on scanned invoices/contracts.
- Extracting tables from scanned PDF's using computer vision algorithms in Python using Microsoft Vision API and Cognitive toolkit.

- Prescriber360 Solutions, Piscataway, New Jersey (Remote)
ML Research Intern, [Feb. 2018 — Mar. 2018]
- Created an End-to-End Recommendation Model for Call Type recommendation personalized for each Customer (Physician).
- Creating Advanced Regression Models to Predict Rx based on Promotional Inputs.