Work Experience

Junior Research Fellow

Indian Institute of Science Education and Research (IISER) Bhopal, Madhya Pradesh

August 2023 – July 2024

At IISER Bhopal, I have been deeply involved in research at the intersection of medical imaging, uncertainty quantification, and deep learning. My work focuses on developing reliable AI-based clinical systems and exploring methods that make predictions more interpretable and trustworthy.

Key Contributions:

• Led research on Mallampati Classification, addressing the complexities of medical image categorisation.

• Conducted extensive analysis on Aleatoric and Epistemic uncertainty, proposing improvements for safer clinical AI.

• Evaluated and benchmarked multiple deep learning architectures including ConvNeXt, AlexNet, EfficientNet, VGG.

• Implemented and tested weighting algorithms to improve model accuracy and F1-score.

• Resolved critical buffer overflow & system optimisation issues, strengthening overall performance.

• Actively contributed to experimentation, debugging, and research documentation.

Software Engineer – Research & Development

Videonetics Technology Pvt. Ltd., Kolkata, West Bengal

August 2021 – July 2023

At Videonetics, I worked on building and optimising real-time AI-driven video analytics systems, with a strong focus on face recognition, multi-GPU optimisation, and secure model deployment.

Key Contributions:

• Customised software installations tailored to client-specific hardware and analytics requirements.

• Improved the performance and stability of the Face Recognition System, reducing recognition time by 40%.

• Designed advanced algorithms & improved model pipelines, increasing accuracy by 2% and reducing data collection time by 60%.

• Integrated GFP-GAN, achieving a 15% improvement in recognition accuracy.

• Introduced multi-GPU support and optimised system throughput on CUDA-enabled devices.

• Implemented NuCypher-based encryption to safeguard deep learning model files and prevent unauthorised access.

• Collaborated with AMD to develop GStreamer plugins for VVAS-based analytics on AI accelerators.

• Facilitated the migration of analytic modules from DeeperLook to the VVAS framework using G-Streamer and Vitis-AI.

Internships

Research Intern - Ciphense AI Research Lab

Bangalore, Karnataka | May 2020 – June 2020

• Built an innovative smart digital whiteboard using Node.js to support real-time online collaboration.

• Developed a smart-draw module that digitises and redraws hand-drawn shapes for cleaner note generation.

• Trained an OCR model using the Bentham dataset to accurately recognise handwritten board content.

• Delivered a complete, functioning prototype enabling teachers to generate and share notes instantly.

Intern - Centre for Development of Advanced Computing (CDAC)

Pune, Maharashtra | Dec 2019 – Jan 2020

• Worked extensively with High-Performance Computing using OpenMP and MPI for parallel & distributed training.

• Built a CNN-powered pneumonia detection web app, reducing diagnosis time by nearly three days.

• Designed a user-friendly interface to upload X-ray images and obtain automated clinical predictions.

• Combined HPC techniques with medical AI to significantly accelerate model training and inference.

Research Intern - National Institute of Technology (NIT)

Rourkela, Odisha | May 2019 – July 2019

• Trained the UCF-Action dataset for accurate human activity recognition in videos.

• Developed an early video caption generation system powered by the trained activity model.

• Applied deep learning to automate video summarisation and content interpretation.

• Proposed a novel extension: generating preliminary police FIR summaries based on crime-scene images.

Research Intern - Indian Institute of Technology (IIT)

Bhubaneswar, Odisha | June 2018 – July 2018

• Gained hands-on experience in ANNs, machine learning, and direct modelling.

• Designed and simulated an ANN-based XOR gate, demonstrating neural network fundamentals.

• Performed Iris flower clustering using k-means as part of an unsupervised learning project.

• Successfully executed independent ML mini-projects, building core foundations in AI.