Raj G.

About Raj G.

Building Co-pilot Discover- Creating agents and infra for Copilot Discover- Worked on Gen AI to add eval metrics on MSN gems to enhance existing quality by 73.33% and safety measures by 33.2%- Agentic AI using Autogen to automate and create a new genre and feel for entire MSN ruby feeds- Working in Microsoft as a full stack on C sharp & React to power the sports cards across windows- Worked as full stack developer in Java, SpringBoot, Spring MVC, Hibernate, JPA, Junits, REST Apis, Microservices from backend, SQL as database and React, JS, TS, Spectrum from frontend perspective in Adobe- Working and managing various projects currently, increasing systems throughput by 45.63% and efficiency by 57.89%- Code Review of peers, managing releases and git branches for deployments of critical sprint releases- Deployment through Jenkins, Docker etc- Ensuring code quality via junits, integration tests, sonar etc- Being a quick learner and a competitive Coder with problems solved on hackerearth, gfg, leetcode, hackerRank etc other platforms, and developed several projects using Spring & Angular - Developed Spring REST apis, familiar with junit testing and branch release deployment in github- Debugged dynamic bugs reported during automation & regression testing- Worked in Agile jira methodologies for consistent progress to meet deadline- Secured 98.71 percentile in Gate CS 2020, having in-depth knowledge of data structure,algorithms, dbms, computer network -I am a topper in Nptel computer organization & published paper in springer on Image Dehazing techniques.Being quick learner & adaptable dynamically I love challenges! Apart from coding,I love puzzles, solving sudokus, playing chess & video games. Never stop watching movies in leisure time though!

Experience

1. Software Engineer 2

   Microsoft

   May 2024 — Present · Hyderabad, IN

   Building Co-pilot Discover - Spearheaded the development of MSN Sports, delivering fullstack functionality with C# and React. Integrated advanced AI systems including Evals and Agentic AI, enhancing the platform’s capability to provide GPT-based judgments for specific prompts on targeted gems- Developed and implemented guideline-based evaluations to ensure the quality and safety of LLM generations. Introduced rule-based evaluations, improving processing efficiency by 22.3%, significantly reducing response time for complex evaluations- Engineered features to fetch grounding data for model training, enhancing the performance and accuracy of predictions. Led the team as DRI (Directly Responsible Individual) to swiftly resolve critical customer issues, driving maximum user satisfaction- Designed and launched multiple high-impact sports and news cards, boosting Daily Active Users (DAUs) through engaging content. Delivered Olympic event cards that garnered significant viewership across MSN, further increasing platform traffic- Deployed code updates and consistently ensured the seamless resolution of customer-facing issues. Played a key role in addressing and mitigating critical security vulnerabilities that posed potential risks for DOS attacks, securing platform stability.

Education

• Ram Mohan Mission High School

  ICSC, Science

• IEM

  Bachelor of Technology - BTech, Computer Science

  2016 — 2020

• The Modern Academy

  ISC