Khagesh Patel

Backend Contributions Across Multiple Teams at Prophecy:

• Metadata Team: Drove significant system performance improvements, leading to enhanced stability, a substantial reduction in operational escalations, and considerable cost savings.
• Custom File Server: Engineered and deployed a custom file server utilizing Netty and zero-copy buffer techniques, resulting in optimized internal service file transfers, reduced memory consumption, and a dramatic improvement in system stability (incidents decreased from 4-5 per month to zero).
• Git Client Optimization: Revamped the server-side Git client to better handle Prophecy's code management, achieving a remarkable 90% reduction in query latency and significantly lowering memory and CPU utilization.
• Transpiler Development: Architected and implemented the parser/lexer and code generation logic for converting Snaplogic and Informatica pipelines into Spark projects, a key initiative that generated $XX million in revenue for the company.
• IDE Sandboxing: Developed and implemented a robust sandboxing mechanism for Python and Scala code during the IDE compilation phase. Further optimized resource utilization by creating a sophisticated sandbox sharing solution to manage diverse dependency requirements across code snippets.
• Copilot AI Features: Played a key role in developing various AI-powered features within the Copilot team, including the generation of dbt SQL projects from user prompts using LLMs.
• Mentorship: Mentored junior engineers on best practices, providing guidance and support to foster their professional growth and career advancement.

• Worked on the buying platform team that is building a new OLAP engine from the ground up.
• Responsible for designing the new OLAP query language and parsing using Antlr
• Implemented validation engine for queries
• Created conversion logic to convert old query to new query
• Built product landing page infrastructure for B2B exporters, I completed the whole backend for this.
• Worked on improving the backend infrastructure of youtube ads to run hotels ads, we scaled it to 200k qps and 100mil arr revenue.

• Played a major part in building products for Cohesity's cloud offerings
• Built the distributed persistence and recovery of AWS VM's using AWS native APIs.
• Contributed in the work on efficient storage and conversion of VMs from HyperV to Azure

• I was involved in building internal applications, for example I created a fast document retrieval system for the team.
• We also performed analysis on internal trading engines to stop unwanted behavior