Namaste! 🙏 Welcome to my portfolio website!
I’m Parth, an AI Data Scientist working on large language models, generative AI, and scalable machine learning systems that help organizations unlock insights from data. 🤖📊
My work focuses on training and fine-tuning LLMs, building high-quality training datasets, and developing evaluation frameworks to improve model reliability, accuracy, and real-world performance.
I’m particularly interested in efficient model adaptation, distributed training, and inference optimization, enabling AI systems to be deployed at scale in production environments. ⚙️🚀
I also work on end-to-end AI pipelines, covering data curation, model training, evaluation, and production deployment for enterprise AI applications. I enjoy collaborating with cross-functional teams to transform research ideas into scalable, production-ready AI solutions.
Ultimately, I’m passionate about building trustworthy and efficient AI systems that empower organizations to make faster, smarter, and more confident data-driven decisions. ✨
💼 AI Data Scientist @ Teradata, Remote (Apr 2026 – Present)
Optimizing LLM/SLM inference with advanced quantization to achieve scalable, low-latency, cost-efficient production AI.
💼 Ph.D. Research Intern @ Samsung R&D Institute India, Bengaluru (Feb 2025 – Nov 2025)
Designed a multimodal foundation model (ALIGN) integrating visual scene features and spatial geometry for location-specific channel generation; led to an IP filing, advancement to the second round of the Samsung Best Paper Award (SBPA).
Built a transformer-style multi-task encoder-decoder (MPC2Vec) for multipath representation, unifying embedding learning, path presence classification, and regression; led to an IP filing.
Developed Vision Transformer-based architectures for spatial AI tasks in coverage prediction, combining scene images with transmitter-conditioned inputs for robust cross-site generalization; led to an IP filing.
Proposed a roadmap for a domain-adapted large language model (ModemInsight) leveraging multimodal telecom log data to automate bug triage, failure analysis, and Q&A in Beyond-5G systems.
Stack used: PyTorch, Hugging Face, Sionna (NVIDIA), OpenCV, Pandas, torchvision, NumPy, Timm, LoRA/QLoRA, LlamaIndex.
💼 Junior Research Fellow @ Indian Institute of Technology Roorkee (Nov 2020 – Aug 2022)
Project: Design and Development of IoT-based Smart Home Automation System (Sponsored by Sikonic Holding Co. Ltd., South Korea)
Designed and developed hardware-software solutions for smart home automation, integrating low-cost sensors, embedded controllers, and IoT connectivity.
Evaluated and compared IoT communication protocols (MQTT, CoAP, HTTP) for low-latency, reliable automation applications.
Built a functional Android application for device control using MIT App Inventor.
Mentored and collaborated with interns to deliver a complete end-to-end prototype for the industry sponsor.
Stack used: C / C++, Arduino IDE, Eclipse Mosquitto, Wireshark, MIT App Inventor.
💼 Summer Intern @ Wipro Consumer Care & Lighting (May 2017 – Jul 2017)
Gained hands-on exposure to lighting production systems and equipment operation, ensuring compliance with safety and quality standards.
Learned fundamentals of industrial process management and equipment maintenance in a large-scale manufacturing environment.
Stack used: SAP ERP, CMMS, Microsoft Excel, SCADA Systems, Six Sigma, ISO Compliance Software.
🎓 (Feb 2021 - Nov 2025) Ph. D in Communication, Network & Signal Processing (ECE)
Indian Institute of Technology Roorkee
Supervisor: Prof. Pyari Mohan Pradhan
Thesis: Information-Theoretic Divergence Measure based Algorithms for Adaptive Filtering and Distributed Estimation.
Developed a novel gradient descent optimization algorithm using Amari-alpha divergence to robustly train machine learning models under error-in-variables (EIV) conditions, ensuring stable parameter learning even when both features and labels are corrupted by noise.
Created a β-divergence-driven loss function and optimization scheme to train machine learning models with high robustness and low complexity, enabling reliable learning under non-Gaussian and heavy-tailed data distributions in real-time AI systems.
Developed divergence-based distributed optimization algorithms using generalized Bregman divergences to enable stable and accurate decentralized training of machine learning models under impulsive noise and adversarial disturbances.
Introduced sparsity-aware and block-sparse distributed learning frameworks that exploit structured sparsity to accelerate convergence and minimize communication, enabling efficient training of large-scale AI models in tasks such as echo cancellation and sparse system identification.
Implemented a communication-efficient distributed training scheme that integrates gradient compression and selective aggregation, reducing bandwidth cost while preserving accuracy in large-scale decentralized AI systems.
Published 6 journal articles and 10 IEEE conference papers on distributed and decentralized AI, with 6 additional journals and 1 patent currently under review.
Stack used: MATLAB, NumPy, Wolfram Mathematica, NetworkX, PyTorch, TensorFlow, padasip, LaTeX, Overleaf, BibTeX, Zotero, Git / GitHub.
🎓 (Aug 2018 - May 2020) M. Tech in Communication Systems & Networks (ECE)
National Institute of Technology Hamirpur
Supervisor: Prof. Rakesh Sharma
Dissertation: Analysis of Machine Learning Algorithms for Brain Tumor Detection.
Designed a Hidden Markov Model (HMM) for brain tumor detection on the BRATS 2013 dataset, published as a book chapter in Springer’s Intelligent Computing and Communication Systems (2021).
Built a deep learning model using U-Net for tumor segmentation on the LGG MRI dataset, achieving strong Dice scores and robustness to variations in tumor shape, size, and location.
Published 1 book chapter on HMM based brain tumor segmentation in Springer Intelligent Computing and Communication Systems.
Stack used: Keras, TensorFlow, NiBabel, Segmentation Models Library, MATLAB, Scikit-learn, Hmmlearn, OpenCV, PyTorch.
🎓 (Aug 2014 - May 2018) B. Tech in Electronics & Communication Engineering
JN Government Engineering College Sundernagar
Major Project: Facial Kepypoint Detection System.
Built a deep learning model using MobileNetV2 with direct coordinate regression on the LaPa dataset for facial keypoint detection, achieving robustness against occlusions, illumination changes, and age diversity.
Stack used: Keras, TensorFlow, Caffe, OpenCV, Dlib, Imgaug, NumPy, Pandas, Matplotlib.
🎓 (Mar 2014) Intermediate
🎓 (Mar 2012) Matriculate
🎯Received best paper award in 2025 IEEE 2nd International Conference on Communication Engineering and Emerging Technologies (ICoCET 2025), Kuala Lumpur, Malaysia.
🎯Submitted an abstract of original research work which was shortlisted for the second round of the prestigious Samsung Best Paper Awards (SBPA). Currently invited to submit the full paper for further evaluation.
🎯Received student travel grant worth $100 from IEEE Communication Society for attending IEEE ANTS 2024.
🎯Received Mukhyamantri Protsahan Yojana Scholarship worth ₹75,000 from Himachal Pradesh Government in 2022.
🎯Secured 94.85 percentile in Graduate Aptitude Test in Engineering (GATE) 2018.
🎯Secured Second Position in Robosprint Event in 2015.
🎯Secured 95.76 percentile in Joint Entrance Examination (JEE) Main 2014.
🎯Cleared BUEST Master Mind-Himachal 2013, Level II competition. Got Entitled for a scholarship worth ₹25,000 in 2014.
(2025) Attentded Workshop on "Unlock the Future - 6G, AI and Security" at Nokia Bengaluru, India.
(2025) Presented paper in "2025 IEEE 2nd International Conference on Communication Engineering and Emerging Technologies (ICoCET 2025)" at World Trade Centre Kuala Lumpur, Malaysia.
(2024) Presented paper and poster in "IEEE International Conference on Advanced Networks and Telecommunications Systems (ANTS)" at IIT Guwahati, India.
(2024) Presented paper in "34th International Conference on Computer Theory and Applications (ICCTA)" at Arab Academy for Science, Technology & Maritime Transport (AASTMT), Alexandria, Egypt.
(2024) Presented paper in "Third International Conference on Power, Control and Computing Technologies (ICPC2T)" at NIT Raipur, India.
(2024) Presented paper in "Third International Conference on Power, Control and Computing Technologies (ICPC2T)" at NIT Raipur, India.
(2023) Presented paper in "7th International Conference on Computer Applications in Electrical Engineering-Recent Advances (CERA)" at IIT Roorkee, India.
(2023) Presented paper in "IEEE International Conference On Electrical, Electronics, Communication and Computers (ELEXCOM)" at IIT Roorkee, India.
(2023) Attended E-Workshop on "Design Challenges of IoT with AI & ML Applications 2.0" at NIT Hamirpur, India.
(2023) Attended "JTG/IEEE Information Theory Society Summer School" at IISc Bangalore, India.
(2022) Presented paper in "IEEE Delhi Section Conference (DELCON)" at NSUT Delhi, India.
(2020) Presented paper in "1st International Conference on Cutting-Edge Technologies in Computing and Communication Engineering (IC4E)" at NIT Kurukshetra, India.
(2019) Attended Faculty Development Programme on "Antenna Trends" at NIT Hamirpur, India.
(2019) Attended Workshop on "Advances in Artificial Intelligence & Machine Learning (AIML-2019)" at NIT Hamirpur, India.
(2016) Attended National Symposium on "Research Innovation in Wireless Technology" at JNGEC Sundernagar, India.
(2016) Participated in Project Gallery during Decibel 2016 at JNGEC Sundernagar, India.
(2024) Python (Basic) - HackerRank
Internship with Sawdah Farooq
Security enhancement in MQTT protocol using QKD.
B. Tech Project (BTP) with Anshul Gusain and Prantaneel Debnath
Secured MQTT based Distributed Tracking over WSN- Hardware Implementation.
M. Tech Thesis with Prem Chand Panwar
Hardware Implementation of Parameter Estimation over WSN.
M. Tech Thesis with Gorla Pavan Kumar Reddy
Kernel Function based Gradient Descent Constrained LMS Algorithm.
B. Tech Lab based Project (LBP) with Anshul Gusain
Hardware Implementation of Consensus Algorithm over WSN.
SPARK Internship with Sahil Singh Balyan
IoT based Home Automation Realization using MQTT.
Internship with Ishan Pandey
Comparison of IoT Protocols for the Application of Home Automation.
IEEE Transactions on Signal Processing.
IEEE Sensors Journal
IEEE Transactions on Circuits and Systems II: Express Briefs.
EURASIP Journal on Wireless Communications and Networking (Springer).
International Journal of Sensors and Sensor Networks.
Signal, Image and Video Processing (Springer).
The World Conference on Computational Science and Technology (WcCST-2026)
International Conference on Smart Technologies and Intelligent Computing (INCSTIC-2025).
IEEE International Conference on Advancement in Computation & Computer Technologies (InCACCT-2025).
International Conference on Artificial Intelligence, Device Computing, Communication, and Signal Processing (AIDCCSP 2024).
IEEE International Conference on Electrical, Electronics, Communication and Computers (ELEXCOM'23).
Adaptive Signal Processing (ECN-614) with Prof. P.M. Pradhan @ IIT Roorkee.
Digital Communication and Signal Processing Techniques (ECN-510) with Prof. P.M. Pradhan @ IIT Roorkee
Digital Communication Laboratory (ECN-510) with Prof. P.M. Pradhan @ IIT Roorkee
Electronics Engineering Lab (EC-101) with Prof. R. Sharma @ NIT Hamirpur.
Digital Signal Processing Lab (EC-317) with Prof. R. Sharma @ NIT Hamirpur.
Institute of Electrical and Electronics Engineers (IEEE). Graduate Student Member (2023 - present).
IEEE Signal Processing Society Member (2024 - present).
IEEE Communications Society Membership (2024 - present).
IEEE Young Professional Member (2023 - present).
Institution of Electronics and Telecommunication Engineers (IETE) Member (2017 - 2019).