Work Experience

• Built and owned the front-end dashboards, defining clear UI boundaries where analytics widgets run as independently loaded and performance-isolated modules, enabling incremental rollout and runtime composition using micro-frontend and Module Federation principles.

• Implemented and deployed the dashboard UI using Next.js App Router on Vercel, leveraging Incremental Static Regeneration to pre-cache dashboard skeletons and achieve sub-500ms First Contentful Paint for data-heavy views.

• Optimized dashboard rendering for 10k+ active records by replacing DOM-heavy SVG charts with Canvas-based D3.js and virtualized tables (TanStack Virtual), and used React Profiler to eliminate unnecessary re-renders and maintain smooth 60fps scrolling.

• Integrated the front-end dashboards with a real-time streaming backend (Node.js, Express, SSE, Kafka consumers) to consume and render high-volume market data with low-latency updates at the Next.js Edge layer.

• Engineered a multi-layer liveness detection pipeline connecting React-based vision modules with Python, Fast API inference endpoints for AI-driven KYC verification and anti-spoofing pipeline using OpenCV, MediaPipe, and DeepFace to detect identity fraud and validate user presence in video feeds

• Integrated WebSocket channels for real-time user feedback and challenge– response flows, achieving p95 end-to-end frame-to-decision latency of 150 ms between client motion tracking and backend validation.

• context-engineered, agent-orchestrated verification pipeline using CrewAI and Vision LLMs to extract and validate critical data from IDs and 30-second video selfies, optimizing finance application workflows. Optimized LLM-based decisioning agents that cross-reference extracted data with employee databases to determine precise loan eligibility and automate approval workflows.

• Containerized and orchestrated distributed microservices using Github and Kubernetes, ensuring multi-AZ resilience, and sustaining a 99.9 % API SLO over a 90-day monitoring window. Automated verification reporting and audit trails with AWS Lambda, S3, and CloudWatch, streamlining compliance workflows and cutting data aggregation time from 4 hours to 1.5 hours (≈ 60 % faster)

• Collaborated with cross-functional teams to organize and lead AI, cloud, and web development workshops, emphasizing practical applications of Google technologies.

• Contributed to a group project, InternEase, by developing the front-end interface using React.js and implementing responsive design systems to enhance usability and scalability

• Engineered front-end prototypes integrating Google APIs such as Maps, Authentication, and Cloud Functions, demonstrating secure and real-time data workflows.

• Designed and implemented an AI-driven customer service chatbot by integrating GPT-3 with a Retrieval-Augmented Generation (RAG) pipeline, enabling context-aware and document-grounded responses for a web-based application.

• Built and deployed 4+ client websites using HTML, CSS, Bootstrap, and JavaScript, collaborating with cross-functional teams to deliver responsive, high-performance, and brand-consistent experiences.

• Led UX/UI redesign efforts for Trogis.com and co-developed Rhabdom.com, improving mobile usability and navigation, resulting in increased session duration and user engagement.

• Built modular React and TypeScript applications & reusable component libraries aligned with Material UI, reducing page latency by 25%, cutting frontend development effort by 30%, and ensuring consistent UI across six core web modules.

• Integrated GraphQL APIs with a Node.js and PostgreSQL backend to support scalable data flows for 10,000+ concurrent users, reduced backend query latency by 20%, and partnered with design and product teams to improve accessibility through lazy loading and dynamic routing, increasing Lighthouse accessibility scores by 18%.

• Automated build and deployment pipelines with Docker, AWS EC2, and GitHub Actions, reducing deployment time by 40% and standardizing continuous delivery across multiple staging environments.

• Implemented comprehensive unit and integration testing using Jest and Cypress, achieving 92% code coverage and minimizing regression defects across major releases.