Research & Industry Experience
Nitte Meenakshi Institute of Technology
Research Assistant | Jan 2021 – Dec 2023
- Designed and developed Trading AutoBot, an AI-driven automated trading system integrating statistical methods, machine learning models, and reinforcement learning–based decision logic for both live and simulated trading environments.
- Architected a modular strategy framework comprising statistical, indicator-based, and ML-driven trading strategies, enabling systematic evaluation and adaptive selection based on asset-specific performance characteristics.
- Developed a large-scale backtesting pipeline using historical market data and the TradersView platform, evaluating strategy robustness across diverse market regimes, volatility conditions, and asset classes.
- Implemented a dynamic strategy-selection algorithm that selects and deploys models from a custom strategy library using performance, risk, and market-behavior metrics, allowing adaptive deployment under changing conditions.
- Integrated brokerage and exchange APIs (Zerodha, Alpaca, Binance) to support real-time and paper-trading execution, ensuring consistency between simulated experiments and live environments.
- Designed and deployed autonomous risk-management and monitoring systems, including position sizing, drawdown control, and event-triggered alerting pipelines to support end-to-end automated operation.
- Led experimentation across multiple market scenarios, analyzing trade-offs between profitability, stability, and risk exposure, with findings contributing to a peer-reviewed IEEE conference publication and an Indian patent filing.
Revoseven
Tech Consultancy | Jul 2023 – Sep 2023
- Developed TruthScore, a prototype multimodal consistency engine for e-commerce platforms, designed to detect discrepancies between product text, images, pricing, and implied user expectations in client-facing systems.
- Designed a novel mismatch-scoring mechanism based on multimodal embeddings and cross-modal similarity constraints, enabling quantitative assessment of semantic and visual consistency across heterogeneous data modalities.
- Built an end-to-end machine learning pipeline covering data ingestion, preprocessing, embedding generation, model training, and cross-modal similarity evaluation for real-world product datasets.
- Addressed practical deployment constraints typical of small and mid-scale e-commerce clients, including limited labeled data, heterogeneous input quality, and the need for interpretable consistency scores.
- Delivered a working prototype to a client as part of a consultancy engagement, demonstrating applicability of multimodal ML techniques in production-oriented business environments.
North Dakota State University
Research Intern | Feb 2024 – Jul 2024
- Designed and implemented distributed versions of evolutionary optimization algorithms including Particle Swarm Optimization (PSO), Glowworm Swarm Optimization (GSO), Genetic Algorithms (GA), and Ant Colony Optimization (ACO) using Apache Spark and HDFS, enabling scalable optimization across synthetic and real-world datasets.
- Architected and deployed a multi-node Spark cluster with HDFS for large-scale experimentation, handling data partitioning, fault tolerance, and workload distribution to support reproducible performance evaluation.
- Executed the first parallel implementation of the Bison classification algorithm, restructuring its computational workflow for distributed execution and validating correctness and performance under varying cluster sizes.
- Proposed one novel PSO-based feature-selection algorithm and implemented two modified PSO variants for high-dimensional datasets, with emphasis on convergence behavior, scalability, and robustness across benchmark data.
- Conducted systematic performance benchmarking using standard optimization test functions such as Schwefel and Ackley, analyzing speedup, parallel efficiency, and convergence trends across increasing data sizes and compute nodes.
- Developed a unified Spark-based experimentation framework integrating classical feature-selection methods and machine learning models, enabling consistent evaluation pipelines using NumPy, Pandas, and scikit-learn.
- Performed comparative analysis of sequential versus distributed implementations, identifying algorithmic bottlenecks, communication overheads, and trade-offs between scalability and solution quality.
- Mentored and technically supervised multiple student research teams as part of the Research Internship Program.
- Documented experimental methodology, results, and system configurations to ensure reproducibility and publication readiness, contributing directly to a peer-reviewed journal article.
Bhoruka Power Corporation Ltd.
Data Engineer | Aug 2024 – Present
- Designed and deployed production-grade analytics systems for energy-generation data across multiple power plants, supporting both operational monitoring and strategic decision-making.
- Built Python-based ETL pipelines to ingest, clean, and integrate multi-source energy and financial datasets, enabling automated data processing and consistent reporting at scale.
- Implemented ensemble machine learning models including Random Forest and XGBoost for energy-demand forecasting and anomaly detection, with emphasis on robustness and interpretability in real-world deployment.
- Optimized SQL queries and data-pipeline performance for high-volume production workloads, reducing reporting latency and improving end-to-end data throughput.
- Developed real-time Power BI dashboards to visualize plant-level performance metrics, forecasting outputs, and anomalies, bridging machine learning outputs with operational decision workflows.
Teaching & Academic Activities
Workshop Instructor & Expected Teaching Assignment
Workshop Instructor — Research & Higher Education Pathways (2025)
- Designed and delivered a one-day academic workshop for undergraduate engineering students focused on research opportunities, research internships, and pathways to postgraduate study.
- Advised students on research project selection, preparation for research internships, and early exposure to academic publishing and graduate-level expectations.
Expected Teaching Assignment — Placement Training Program (Starting Jan 2026)
- Confirmed to teach core undergraduate subjects including Data Structures & Algorithms, Database Management Systems, Object-Oriented Programming, and programming fundamentals in Python and C++ to pre-final-year students.
- Responsible for conceptual instruction, problem-solving sessions, and technical skill development aligned with industry and academic expectations.