Highly accomplished AI/ML Research Lead and Data Scientist with a strong academic background (M.Sc. in Data Science) and extensive experience in developing advanced AI/ML systems, particularly in LLMs, network security, and distributed systems. Excels at leading research and development initiatives, translating complex data into actionable insights, and driving innovative solutions that significantly enhance system performance, security, and compliance. Expertise spans large-scale data analysis, quantum computing applications, and open-source contributions, positioning him as a valuable asset for cutting-edge technical leadership roles.
Work
India Internet Foundation
|
Research Fellow - Research & Development
Kolkata, West Bengal, India
→
Summary
Led the development of OLAP pipelines for large-scale DNS data analysis and a nation-scale distributed DNS system, enhancing performance and security.
Highlights
Spearheaded OLAP pipeline development, enabling analysis of ~12TB of raw DNS data across 700 million domains in collaboration with Internet Society and ICANN.
Directed the development of a nation-scale, distributed DNS-in-a-Box project, significantly improving performance and security for critical internet infrastructure.
India Internet Foundation
|
Research Associate - Research & Development
Kolkata, West Bengal, India
→
Summary
Implemented IPv6 protocol performance metrics and maintained a nationwide distributed measurement network to ensure continuous operation.
Highlights
Implemented IPv6 protocol performance and diagnostic metrics, aligning with RFC8250 standards and integrating them into the existing national measurement network.
Ensured continuous operation and reliability of a nationwide distributed measurement network, supporting critical infrastructure monitoring.
Zeron (TeamCognito Solutions Pvt. Ltd.)
|
Research Lead - Research & Development
Kolkata, West Bengal, India
→
Summary
Led data collection, analysis, and system development for threat analytics, enhancing visibility and actionable insights for network security.
Highlights
Collected and analyzed vast volumes of log data from 150+ network devices, leveraging OLAP pipelines to provide full system visibility and generate actionable threat insights.
Directed the end-to-end development of an External Attack Surface Management system, encompassing scanning, data ingestion, ETL, and dashboarding functionalities.
Significantly contributed to the "risk analysis" (QBER) system, a core component of the Zeron Platform, enhancing its predictive capabilities.
Zeron (TeamCognito Solutions Pvt. Ltd.)
|
Backend Lead - Product Engineering
Kolkata, West Bengal, India
→
Summary
Co-designed and implemented the internal architecture for the core product "Zeron", optimizing backend performance and data access patterns.
Highlights
Co-designed and implemented the internal architecture for the core "Zeron" product, optimizing backend performance, data warehousing, and data access patterns.
Managed the integration of Python-based microservices and AWS cloud infrastructure to support scalable product operations.
Enhanced data access patterns and system design for improved efficiency in data retrieval and processing.
TeamCognito Solutions Pvt. Ltd.
|
Machine Learning Lead - Technical Team
Kolkata, West Bengal, India
→
Summary
Delivered over 3 machine learning-based analytical systems, ensuring flawless operation and high data integrity for technical solutions.
Highlights
Successfully delivered over 3 machine learning-based analytical systems, ensuring flawless operation and high data integrity.
Applied Python, data modeling, and data visualization techniques to develop robust analytical solutions.
Utilized ETL processes, Big Data technologies, and AWS infrastructure to support scalable ML deployments.
Education
St. Xavier's University, Kolkata
Kolkata, West Bengal, India
→
M.Sc.Computer Science (Specialization in Data Science)Grade: 7.69/10 CGPA
Kazi Nazrul University
Kolkata, West Bengal, India
→
B.Sc.Computer ScienceGrade: 8.3/10 CGPA
Publications
Fast classical simulation of qubit-qudit hybrid systems
IEEE Software
Presented a fast classical simulation method for qubit-qudit hybrid systems, advancing the understanding of quantum computing architectures.
Proof of Fair-Chance: An Unbiased Approach to Internet Voting using Fair Chance in Proof of Stake
2025 International Conference on Artificial Intelligence & Sustainable Computing (AISC)
Presented a novel approach to internet voting, ensuring fairness and unbiased outcomes through a fair chance in Proof of Stake mechanism.
A Layered Blockchain Architecture for Unbiased and Trustworthy Internet Voting Systems
2025 IEEE 6th India Council International Subsections Conference (INDISCON)
Proposed a layered blockchain architecture designed to enhance the trustworthiness and impartiality of internet voting systems.
Quantum Circuit Synthesis of an Approximate Hybrid Kolmogorov-Arnold Network
IEEE Computer Society Annual Symposium on VLSI (ISVLSI)
Explored the synthesis of quantum circuits for an approximate hybrid Kolmogorov-Arnold Network, contributing to quantum computing advancements.
A near-term quantum simulation of the transverse field Ising model hints at glassy dynamics
The European Physical Journal Special Topics
Investigated near-term quantum simulation of the transverse field Ising model, revealing insights into glassy dynamics.
Approximated Approaches to the Vehicle Routing Problem by Graph Reduction
IEEE Access
Developed approximated approaches for the Vehicle Routing Problem utilizing graph reduction techniques to improve efficiency.
Code Generation from Natural Language Statement using Self-Repairing based T5 Model
St. Xavier's University, Kolkata
Developed a T5 model incorporating self-repair mechanisms for accurate code generation from natural language statements.
Pre-Quantum to Post-Quantum Cryptography Transition: A Journey Connecting the Security and Challenges Eras
Integration of AI, Quantum Computing, and Semiconductor Technology. IGI Global
Explored the transition from pre-quantum to post-quantum cryptography, addressing security challenges and future directions.
FragQC: An efficient quantum error reduction technique using quantum circuit fragmentation
Journal of Systems and Software, 214, 112085
Introduced FragQC, an efficient quantum error reduction technique based on quantum circuit fragmentation.
QuDiet: A Qubit-Qudit Hybrid Quantum Simulator with Benchmark Circuits
IET Quantum Communications
Developed QuDiet, a qubit-qudit hybrid quantum simulator, and evaluated its performance with benchmark circuits.
Qurzon: A Prototype for a Divide and Conquer-Based Quantum Compiler for Distributed Quantum Systems
SN Computer Science, 3(4), pp.1-14
Presented Qurzon, a prototype for a divide-and-conquer-based quantum compiler designed for distributed quantum systems.
Languages
English
Hindi
Bengali
Skills
Artificial Intelligence & Machine Learning
LLMs, Natural Language Processing (NLP), Machine Learning, Deep Learning, Anomaly Detection, Genetic Algorithms, Quantum Computing, Statistical Analysis, Data Modeling, Data Visualization.
Programming & Development
Python, SQL, Linux Kernel Driver, Event Driven Architecture, Microservices.
Cloud & DevOps
AWS, ETL, Airflow, Prefect, Tool Automation.
Data Engineering & Big Data
OLAP, DNS, DNSSEC, Clickhouse, Big Data, Data Warehouse, Data Ingestion.
System Design & Architecture
Distributed System Design, System Design, Optimization, IPv6.
Built a developer-friendly yet fast DNS resolution library for Python.
Arxiv Summarizer
→
Summary
Co-developed a Python library for automated fetching, parsing, and summarizing scientific documents from arXiv, streamlining research workflows.
Code Generation from Natural Language Statement using Self-Repairing based T5 Model
→
Summary
Designed and implemented an LLM model for code generation, integrating a 'self-repair' mechanism to enhance accuracy and reliability for complex coding tasks. This project served as a major component of a Master's thesis.
Natural Language to SQL
→
Summary
Engineered an NLP-based system to translate natural language queries into structured SQL, integrating with a data warehouse and supporting over 20 distinct features using open-source LLM models.
Compliance Control Mapping LLM
→
Summary
Led the development of an innovative compliance auditing system, leveraging LLMs and a parallel rule engine to automatically map compliance controls from complex PDF documents to specific regulations and standards, enhancing audit efficiency.
Analyzed and optimized an existing Quantum Simulator by leveraging distributed and high-performance computing techniques, improving simulation efficiency.
Research Intern Project: Genetic Algorithm for Circuit Fragmentation
→
Summary
Developed a Genetic Algorithm-based circuit fragmentation algorithm, enhancing circuit cutting strategies for complex systems.
Research Intern Project: Vehicle Routing Problem on NISQ Devices
→
Summary
Investigated and applied approximation approaches to solve the Vehicle Routing Problem on NISQ Devices, utilizing classical preprocessing techniques.
PyAttck Contribution
→
Summary
Contributed a fix to PyAttck for compatibility with the updated MITRE data format.
Network Log Anomaly Detection
→
Summary
Developed and analyzed a real-time anomaly detection system for network and system logs, utilizing Granger causality to identify causal relations for advanced threat analytics.
opensearch-py Contribution
→
Summary
Contributed support for the alerting plugin in opensearch with its low-level API for Python.
Insurance Claim Authenticity Classification
→
Summary
Directed the development and deployment of a full end-to-end ML system, employing a hybrid Decision Tree and Deep Learning approach to accurately classify insurance claim authenticity.
TensorFlow.NET Core Routines Contribution
→
Summary
Contributed to the development of core routines for TensorFlow.NET.