Ankit Kumar

Designed, built, and deployed robust distributed Python data pipelines, seamlessly integrating AI model outputs with a React.js frontend via REST APIs, thereby decoupling presentation and model layers for enhanced maintainability and scalability in web-based applications. Architected scalable FastAPI microservices with strict OpenAPI/Swagger contracts, facilitating independent deployment and robust integration of backend AI services with frontend components, adhering to best practices for web development frameworks. Configured and optimized GitHub Actions CI/CD pipelines from scratch, automating test and deployment workflows to streamline collaborative development and ensure rapid, reliable delivery of web application updates.

AI-Integrated Web Development Intern at IISER Berhampur (2025-05 – 2025-07)

• Designed, built, and deployed robust distributed Python data pipelines, seamlessly integrating AI model outputs with a React.js frontend via REST APIs, thereby decoupling presentation and model layers for enhanced maintainability and scalability in web-based applications.
• Architected scalable FastAPI microservices with strict OpenAPI/Swagger contracts, facilitating independent deployment and robust integration of backend AI services with frontend components, adhering to best practices for web development frameworks.
• Configured and optimized GitHub Actions CI/CD pipelines from scratch, automating test and deployment workflows to streamline collaborative development and ensure rapid, reliable delivery of web application updates.
• Introduced structured logging across all web-based AI services, significantly reducing diagnostic investigation time by resolving recurring service failures and improving system observability for performance optimization.

Computer Vision Engineering Intern at NIST University (2024-07 – 2024-08)

• Deployed a production-grade Computer Vision classifier to edge hardware, rigorously optimizing performance under strict memory and compute constraints to achieve reliable, low-latency inference in a real-world environment.
• Automated a manual model evaluation pipeline in Python, boosting operational efficiency by achieving a 3x speedup and recovering over 6 engineering hours weekly, contributing to faster iteration cycles for web-related content updates.
• Profiled and resolved recurring Linux system failures impacting deployed ML services, implementing comprehensive production monitoring to proactively track model accuracy and ensure continuous, optimized service performance.

B.Tech in Computer Science & Engineering – NIST University (2022-08 – 2026-04)