Hello INCF Community,
My name is Parth Bhatnagar, and I am an AI Researcher and Data Science professional currently pursuing my undergraduate degree at Manipal Institute of Technology. I am deeply passionate about building scalable, research-driven intelligent systems that bridge advanced AI with real-world scientific impact — particularly in domains where open science, reproducibility, and data-driven discovery are critical.
Over the past few years, I have worked with organizations such as Intel, IBM, Infosys, and Bosch, contributing to AI-driven systems across multimodal learning, document intelligence, predictive analytics, and large-scale data processing. At Intel, I developed an intelligent document analysis pipeline using LayoutLMv3, OCR, NER, knowledge graphs (Neo4j), and semantic search, achieving ~92% layout detection accuracy. I also built multimodal conversational systems leveraging Whisper, FAISS, and LLM-based QA — integrating structured and unstructured data into unified knowledge systems.
Research, Patents & Publications
Research forms the foundation of my work:
- Author of 25+ research papers published in IEEE, ACM, Springer, and Elsevier venues
- Publications in reputed conferences such as COMSNETS, PREMI, ICMLT, ICISS, and ACMLC
- 3 Q1/Q2 indexed journal publications
- Holder of 7+ Indian copyrights/patents
- Author of 4 technical books on AI, Software Engineering, Cybersecurity, and Cloud Computing
- Recipient of $7500 research funding from IBM
- Invited multiple times as a guest speaker at engineering institutions
- Recognized 3 times as a LinkedIn Top Voice
For me, research is not limited to publication — I focus on building deployable systems grounded in scientific rigor.
My projects include Federated Learning for patient mortality prediction using 60,000+ critical care records, GAN-based image super-resolution trained on 100,000 images, RAG-powered conversational AI systems, and multimodal visual question answering models. I have hands-on experience working with large-scale datasets, distributed systems, model reproducibility, and performance evaluation — aspects that strongly resonate with INCF’s emphasis on neuroscience data standards, FAIR principles, and interoperable research tools.
Technically, I work extensively with Python, PyTorch, TensorFlow, Spark, Hadoop, Docker, Kubernetes, and cloud platforms like AWS and Azure. I enjoy designing end-to-end AI pipelines — from data engineering and preprocessing to model training, validation, and deployment — while ensuring scalability and reproducibility.
What excites me about INCF is its commitment to advancing open neuroscience through standards, infrastructure, and collaborative tools. I am particularly interested in contributing to projects involving scientific data pipelines, knowledge integration, multimodal data analysis, reproducible ML workflows, or AI systems that can support neuroscience research communities globally.
I am eager to collaborate, contribute meaningfully, and grow within the INCF open-science ecosystem during GSoC 2026.
Looking forward to connecting and contributing!
I am Stefan, 27, M. Sc. in