Responsibilities included:

• Contributed to rewriting existing Node.js microservices in Rust using async/await, helping the team reduce P99 latency by 50% and cut infrastructure costs by 20% after deployment
• Engineered a concurrent client-to-S3 multi-part upload API utilizing parallel chunk processing algorithms, slashing large-file transfer latency by 83% (4 min to 40 sec) and optimizing throughput.
• Developed parts of a gRPC-based backend in Dart, enabling bi-directional streaming and contributing to a 40% reduction in API payload size compared to previous REST endpoints.
• Worked on an internal cleanup module that integrated Nest.js APIs with Rust workers, supporting the removal of 7,000+ orphaned documents and helping reduce database storage usage by 30%.
• Assisted in building S3-backed ingestion scripts for ML workflows, improving data availability for model training and contributing to a 15% reduction in training time through more efficient data handling.

Responsibilities included:

• Built individual ERP modules using Next.js and PostgreSQL, improving workflow organization across manufacturing teams.
• Implemented RBAC and added targeted query indexes, contributing to a 20% improvement in API response times for 1,000+ concurrent active users.
• Wrote 30+ PostgreSQL triggers to automate routine supply-chain steps, reducing manual work by 50%.

Responsibilities included:

• Implementing the optimal ways to use reinforcement learning to identify high-risk patient conditions and suggest the best treatments in healthcare, while also improving clinical decision-making systems to provide better, personalized care for patients.
• Implemented WGAN-based data augmentation for limited EEG datasets under faculty supervision, improving motor-imagery classification accuracy to 72%.
• Fine-tuned transformer NLP models for legal-judgment prediction (achieving 82% accuracy) and helped build ETL pipelines using GPT, LangChain, Docker, and AWS for scalable experiments.
• Developed a Q-learning prototype for adaptive traffic-signal control in simulation, showing a 40% congestion reduction and 15% lower wait times
• Automated Python scripts for web-metric analysis, increasing evaluation throughput by 30% and improving dataset reliability by 8%.

Responsibilities include:

• Designed and implemented end-to-end machine learning pipelines, including data preprocessing, feature engineering, model training, validation, and deployment.
• Collaborated with clients from various countries to understand project requirements, define objectives, and deliver tailored machine learning solutions.
• Worked on Computer Vision, GANs and various Deep Learning Projects showcasing expertise in these domains.

Responsibilities include:

• Developed ISPSO, an optimized image segmentation method using meta-heuristic algorithms, achieving 80% segmentation accuracy. This approach leverages Particle Swarm Optimization (PSO) for enhanced precision in medical and industrial image analysis.
• Built a football match prediction model incorporating Kelly Criterion to optimize betting strategies, achieving 75% accuracy. Analyzed historical match data, odds, and team performance using machine learning for data-driven betting decisions.

Responsibilities include:

• Designed and implemented a robust Computer Vision Algorithm, leveraging OCR Tools such as PaddleOCR & EasyOCR.
• Spearheaded the design and implementation of a mathematical algorithm, resulting in a 20% improvement in system performance.
• This achievement contributed significantly to project enhancements and played a pivotal role in boosting overall operational efficiency in our research.

Responsibilities include:

• Spearheaded a project team of 3 members, overseeing the successful completion of tasks and ensuring alignment with strategic goals.
• Conducted an exhaustive analysis of the Electric Vehicle Market in India, utilizing advanced segmentation techniques and generating a predictive model.