Peenal Gupta

Building the Future with AI & Data Science

Applied Data Scientist | Co-Founder at PinakashieldTech | Seeking PhD positions | Cybersecurity Innovator.

Bridging the gap between software architecture and advanced machine learning.

Deep Learning Distributed Systems Trustworthy AI Privacy Benchmarking LLMs Intrusion Detection Generative AI LLMs Federated Learning Cybersecurity
Get in Touch

Entrepreneurship & Vision

Co-Founder

PinakashieldTech OÜ | Estonia | 2025 - Present

Founded a cybersecurity and AI company aligned with EU digital sovereignty.

PinakashieldTech is a next-generation cybersecurity firm specializing in AI-driven, privacy-first solutions for IT-OT convergence. We bridge the gap between cloud security and critical infrastructure by deploying quantum-safe identity management and federated threat intelligence. Our core platform leverages generative AI for proactive vulnerability assessment and autonomous recovery, ensuring resilient operations against evolving quantum-era threats.

Key Focus: IT-OT Security, Federated Learning, Quantum-Safe Cryptography, and Autonomous Threat Defense.

Education

M.Sc. Applied Data Science & Analytics

SRH University Heidelberg, Germany
2023 - 2025

Focused on applied machine learning, statistical analysis, and data-driven decision making.

B.Tech. Computer Science and Engineering

Rajasthan Technical University, India
2008 - 2012

Master's Thesis

I-Driven Chemical Research: Benchmarking Large Language Models for Qualitative and Quantitative Analysis
Advisor: Dr. Niklas Wulkow | BASF SE, Ludwigshafen Germany | 2024 - 2025

Objective: Develop and evaluate LLM-based frameworks for automatic extraction and structuring of chemical experiment data from scientific literature.

Key Contributions:

  • Designed domain-specific evaluation metrics for assessing LLM performance in chemistry domain
  • Built Python pipelines integrating GPT-4, LangChain, and Azure OpenAI for structured data extraction
  • Evaluated multiple LLMs (GPT-4, Claude, Mistral) on chemical reaction data extraction tasks
  • Created benchmarking framework to compare performance across different model architectures

Technologies: Python, LLMs (GPT-4, Claude), LangChain, Azure OpenAI API, Prompt Engineering, Evaluation Metrics

Impact: Framework enables automation of chemical data extraction, reducing manual effort for BASF researchers and accelerating computational chemistry workflows.

Publications

Federated Learning-Driven Intrusion Detection for Cybersecurity in Smart Distribution Systems
IEEE GLOBECOM Workshops, 2024 | DOI

Proposed a federated transfer learning approach for smart grids to enable privacy-preserving DDoS/MITM detection.

A privacy-preserving framework for mitigating network attacks in distributed systems using FedGen-AI
International Journal of Pervasive Computing and Communications | DOI

This study aims to address critical security gaps in distributed systems by developing and evaluating a novel framework that integrates Federated Transfer Learning (FTL), Generative AI (GenAI) and Blockchain technology. Research specifically enhances the detection and mitigation of sophisticated network attacks, including distributed denial of service, man-in-the-middle and model poisoning.

Introduction to Number Theoretic Transform (NTT)
arXiv preprint, 2025 | arXiv

Explored NTT as an FFT-like tool over finite fields, analyzing cyclic convolution and fast NTT/INTT algorithms.

Revisit to the Bai-Galbraith signature scheme
arXiv preprint, 2025 | arXiv

Analysis of lattice-based signature schemes based on Learning with Errors (LWE), contrasting with Dilithium.

Professional Journey

Data Analyst (Master Thesis | Working Student)

BASF SE | Ludwigshafen, Germany
Nov 2024 - Sept 2025
  • Designed Python pipeline for extracting chemical data using GPT-4, LangChain, and Azure.
  • Created novel evaluation metrics for benchmarking LLMs in chemical research.

Research Intern (HPC)

Jülich Supercomputing Centre | Germany
Aug 2024 - Sept 2024
  • Benchmarked astrophysics N-body integrators on pre-exascale systems (JURECA).
  • Applied MPI, OpenMP, and CUDA for HPC optimization.

Research Associate

Hochschule Worms | Germany
Dec 2023 - Jun 2024
  • Built anomaly detection models using Bayesian Deep RL + CNNs.
  • Implemented real-time dashboards with TimescaleDB.

Product Delivery Lead

The Adecco Group | Lyon, France
Jun 2022 - Dec 2022
  • Owned end-to-end product lifecycle and coordinated cross-functional squads.
  • Managed Agile workflows using Azure DevOps/GitLab.

Technical Lead & Scrum Master

Infosys | Lyon, France
Sep 2019 - Jan 2022
  • Led 10 devs + 4 QA in transportation domain.
  • Enforced code quality and facilitated full Scrum ceremonies.

Technology Analyst

Infosys | Pune, India
May 2016 - Aug 2019
  • Full-stack microservices for new client apps; REST APIs with .NET Core, Angular 4/5, TypeScript; TDD with xUnit.
  • Additional retail track: Sitecore-based sites for a global wholesale group.

Software Engineer

HCL Technologies | Noida, India
Oct 2013 - May 2016
  • Built C#/WCF services for MSN browser markets; delivered a Windows mobile app (GSG) for marketer workflows.
  • Web development across Microsoft programs (SharePoint/Sitecore); HTML/CSS/JS/jQuery.

Web Developer

Micronic Infotech | Ajmer, India
Jul 2012 - Sep 2013
  • End-to-end PHP website builds and deployments (PHP/HTML/CSS/JS).

Technical Expertise

Programming

Python R C# TypeScript SQL

AI & HPC

PyTorch TensorFlow Federated Learning MPI/OpenMP CUDA

Cloud & Data

Azure Google Cloud Docker Postgres TimescaleDB

Get In Touch

Interested in collaboration or my research? Leave a message below.