My core responsibilities revolves around architecting, developing and maintaining the UI for Yellow’s no-code conversational AI platform.

• Architected and developed a no-code workflow builder for support agents, enabling the creation of 15k+ workflows to date, and saving over 1,000+ agent hours.

  📄 Documentation

• Created an innovative flow-debugger tool, similar to debuggers in IDEs. This helped bot developers in the quick identification of logic and runtime errors, thus reducing debugging time by 20-30%.

  📄 Documentation

• Developed a new API management node which simplified the process of integrating external APIs in flows. It had advanced features to import JSON/cURL, and configuring a response transformation function. Reduced context switching by around 70% (from 3-4 screens to just 1).

  📄 Documentation

• Architected and developed a new product offering from scratch called Dynamic Conversation Designer, a no-code tool for designing smart conversational experiences. This feature’s launch increased user sign-ups 2X and was also featured in Forbes.

  📄 Documentation

• Improved the login page’s load time by boosting the Lighthouse score by 32%, raising it from 53 to 78. Also implemented optimization and code-splitting strategies, which resulted in a 66% reduction in transfer size (from 1.5 MB to 500 KB).

• Developed an in-house WYISWYG SSML editor using DraftJS which reduced the effort of writing complex SSML for voice bots and increased adoption for building voice bots.

• Delivered multiple other features end-to-end within high-paced timelines by ensuring close collaboration with stakeholders, product, and design teams—managing the entire lifecycle from requirement gathering and implementation to documentation and successful deployment.

• Proactively contributed to the holistic growth of the UI team by establishing comprehensive code-review guidelines and standards, delivering tech talks, developing a reusable design system library, implementing tools to enhance developer experience, participating in hiring processes, and mentoring junior team members.

My core responsibilities revolved around developing and maintaining the UI for the company’s flagship Credit Underwriting and ML engineering products.

• Developed a Drag and Drop Credit decision authoring tool using JointJS which helped Credit Underwriters configure custom decision flows resulting in reducing the authoring time by 5x and enabling complex decision graphs for approving/rejecting loans.

• Developed a SQL Query generation utility which increased adoption of novice users to perform different kinds of SQL transformation and joins on a dataset with ease.

• Developed a customisable Credit Quality Analysis dashboard using PlotlyJS to visualize and analyze the quality of decisions made by the Decision Engine.

• Developed an interactive Graph-based visualization tool for ML Modelling Workbench to visualize and identify co-relations in a dataset.

• Gracefully handled performance bottlenecks while rendering large sets of data by using techniques like Virtualization and Pagination.

• Was actively involved in the development and improvement of the company’s user-facing website built using Next.js and performed optimization based on reports generated from GTMetrix
• Integrated Google Analytics to analyze a visitor’s flow and helped reduce the drop-off rate by optimizing critical assets and faster load times.
• Worked on the UI development of an AI Analytics Platform which involved connecting to data pipelines and handling large sets of data.