Oct 2024 – Present · 1 yr 8 mos · Bengaluru · On-site

• Developed a Agentic AI Chat application integrated inside a data driven dashboard, allowing users to access data, ask questions, and support users. Utilized React for frontend with websocket for continuous communication and Node.js backend microservice. Uses http based Model Context Protocol with OpenAI LLM, Redis for session management and BullMQ for asynchronous tasks like saving data in MySQL, updating credits and performing evaluations. Subscription based model with Stripe payment integration and previous chat management.
• Allow users to see visual charts, graphs, download report, excels sheets in the chat.
• Implemented a dynamic white-labelling system to render brand-specific interfaces across the platform, empowering clients to customise their dashboards and contributing to over $191K in new revenue opportunities.
• Scaled the infrastructure to support a 4× increase in client data volume, upgrading the Azure Database for MySQL with higher vCore capacity and implementing read replicas to enhance query performance and handle increased data load. Vertically and horizontally scaled virtual machines for critical workloads including API services, data extraction, and background processing, ensuring consistent performance under heavier demand. Expanded storage capacity using tiered storage optimization and moved 6-month-old data to cold storage, significantly reducing costs while efficiently managing larger datasets. Implemented Grafana dashboards to monitor system performance, resource usage, and service health, enabling proactive scaling and faster issue detection.
• Built an evidence-fetching microservice to retrieve, merge, and store video clips from Ring Camera platform for order dispute evidence, streamlining support workflows with automated video handling and delivery.