Satyaki – AWS, Python, Terraform

This project involved developing a comprehensive, end-to-end web application with a frontend, backend, AI, and database components. The product can be plugged into a Sprinklr environment — it's a focused add-on that provides users with exactly the extra capability they need, without requiring the onboarding of another software.

• Led a team of 3 people, including designers and junior software developers.
• Was handling the product owner role, gathering requirements, converting business requests to technical requirements, and communicating with the team.
• Created cloud infrastructure, built CI-CD pipelines.
• Provided code reviews (Python and JavaScript code).
• Carried out the implementation of Claude AI.
• Integrated payments.

One of the first metal poster-selling websites in India. As part of the mini-launch, the client was looking for a proficient full-stack developer to handle the complete lifecycle, from building to production. This website now handles about 300 visitors daily.

• Took on the task of leading a small team of 1 UX designer, 1 app developer, and myself to get this product ready.
• Created frontend and backend architecture.
• Built frontend and backend.
• Integrated an AI chatbot.
• Implemented SEO.

This project involved designing a pipeline that processes 80 GB of image data per day. The pipeline would ingest, clean, add metadata, convert the data to tabular format, and store it in a relational database. The complete infrastructure was designed using Terraform and deployed in development, testing, and production environments. Application teams successfully used the infrastructure to deploy proprietary image processing logic.

• Deployed the pipeline with 96% uptime.
• Trained juniors in writing infrastructure as code.
• Handheld application teams to use this new infrastructure, substituted with proper documentation.
• Debugged user-specific use cases while adding them as new features.

A Dutch information technology company that specializes in using crowd-sourcing for business in the retail, tech, and care sectors. The project focused on the full-stack DevOps pipeline.

• Built an end-to-end data scraping pipeline using Docker and AWS components to scrape multiple websites daily.
• Wrote Python pipelines for wrangling this data - cleaning, Spark processing and storage - all via Airflow orchestration and event-driven.
• Implemented horizontal scaling on each node.
• Orchestrated complete pipeline using Airflow.