Oracle E-Business Suite - Payroll
Integrated payroll application enhancements according to legislative changes, benefiting 100s of customers and 1M+ users. Built web dashboards and online reports using Oracle Application Framework.
Portfolio — Always Learning
Engineering agentic AI and distributed systems at enterprise scale.
Every section a star — born, burning, collapsing. Always learning.
Senior Software Engineer specializing in agentic AI systems and large-scale distributed platforms. I led the Agentforce AI agent platform at Salesforce — adopted by 60,000+ employees with ~$4M in savings — and now lead agent-to-agent (A2A) orchestration. I build LLM-powered agents and real-time event systems running in production at enterprise scale, across Java microservices, Kafka, and AWS.
Oracle → GlobalFoundries → Salesforce — building systems that scale.
A career measured in the things it moved.
Burning brightest now: leading AI agent work that became a guiding star at Salesforce.
I led the LLM-powered AI agents on Salesforce Agentforce — RAG-grounded tools over enterprise knowledge and an offline/online evaluation system — and now lead agent-to-agent (A2A) orchestration: intent routing, cross-agent memory, MCP tool servers, and HITL guardrails running safely in production.
Agentforce / Agent-to-Agent (A2A) / LangGraph / MCP / RAG / HITL
Integrated payroll application enhancements according to legislative changes, benefiting 100s of customers and 1M+ users. Built web dashboards and online reports using Oracle Application Framework.
Developed a highly scalable mobile and web application for tracking real-time user locations with background operation, historical data storage, and Google Maps integration.
Built an application to check semiconductor chip quality through complex rules, saving $2-3 Million per year. Also developed mission-critical engineering change request system.
Used Locality Sensitive Hashing to map similar documents into same buckets. Implemented Jaccard and Cosine similarity measures with BloomFilter for efficient storage.
Built web graphs from crawler data, calculated page rank values, and analyzed node visitation patterns across web graph structures.
Implemented a remote procedure call framework using system calls for client-server communication over UDP sockets.
Designed a distributed file system allowing multiple clients to remotely access files with a local-like experience through caching.
Developed optimized indexing algorithms integrated with Apache Lucene for efficient data retrieval and search across document collections.
Built a unified API for data scraping from prominent websites like Wikipedia, Google (Auto Complete and Trends), Twitter, and Social Mentions for social scientists and data analysts.
Parsed political ads on YouTube to categorize content, identify speakers, present ad stance (attacking or defending), and recognize political affiliations using ML techniques.
Published multiple Android applications on the Google Play Store, demonstrating mobile development expertise.
Published BlackBerry applications on BB App World. Won a BlackBerry Z10 and PlayBook Tablet in the BB10 Apps Challenge.
Languages, agent frameworks, and the platforms in between.
Open to hard problems in AI agents and distributed systems.