Hi, I'm Anirudh Sivakumar
A passionate Computer Science graduate from UC Santa Cruz specializing in LLM logic fine tuning, full-stack development and machine learning. I build immersive web experiences and intelligent systems that solve real-world problems.
Experience
Gen AI Technical Advisor Intern
Key Achievements
- NDA :(
Projects
Built AI-powered Discord bot using Gemini API and Go, serving 800+ users with smart moderation. Implemented gRPC microservices on Cloud Run with Pub/Sub for real-time analytics.
- Serving 800+ users with smart moderation
- 99.7% uptime using Google Cloud auto-scaling
- Real-time analytics with Pub/Sub
Engineered GraphRAG architecture with Neo4j knowledge graphs and Pinecone vector database. Achieved 95% search accuracy processing FDA Orange Book with 50,000+ medication entries.
- 95% search accuracy with 50,000+ medication entries
- Reduced medication cost discovery time by 80%
- Real-time savings calculations
Developed RAG AI chatbot processing 50,000+ data points from 300+ courses with 95% accuracy. Reduced search time by 60% for 1000+ daily queries using Anthropic API and vector database.
- 95% accuracy with 50,000+ data points
- 60% faster search time reduction
- 1000+ daily queries processed
Built AI voice agent using VAPI API to bypass automated menus and connect with human representatives. Improved voice agent response latency by 70% implementing Groq's inference engine.
- Improved voice agent response latency by 70%
- Implemented Groq's inference engine
- Automated menu navigation system
Built AI-powered Discord bot using Gemini API and Go, serving 800+ users with smart moderation. Implemented gRPC microservices on Cloud Run with Pub/Sub for real-time analytics.
- Serving 800+ users with smart moderation
- 99.7% uptime using Google Cloud auto-scaling
- Real-time analytics with Pub/Sub
Engineered GraphRAG architecture with Neo4j knowledge graphs and Pinecone vector database. Achieved 95% search accuracy processing FDA Orange Book with 50,000+ medication entries.
- 95% search accuracy with 50,000+ medication entries
- Reduced medication cost discovery time by 80%
- Real-time savings calculations
Developed RAG AI chatbot processing 50,000+ data points from 300+ courses with 95% accuracy. Reduced search time by 60% for 1000+ daily queries using Anthropic API and vector database.
- 95% accuracy with 50,000+ data points
- 60% faster search time reduction
- 1000+ daily queries processed
Built AI voice agent using VAPI API to bypass automated menus and connect with human representatives. Improved voice agent response latency by 70% implementing Groq's inference engine.
- Improved voice agent response latency by 70%
- Implemented Groq's inference engine
- Automated menu navigation system
Built AI-powered Discord bot using Gemini API and Go, serving 800+ users with smart moderation. Implemented gRPC microservices on Cloud Run with Pub/Sub for real-time analytics.
- Serving 800+ users with smart moderation
- 99.7% uptime using Google Cloud auto-scaling
- Real-time analytics with Pub/Sub
Engineered GraphRAG architecture with Neo4j knowledge graphs and Pinecone vector database. Achieved 95% search accuracy processing FDA Orange Book with 50,000+ medication entries.
- 95% search accuracy with 50,000+ medication entries
- Reduced medication cost discovery time by 80%
- Real-time savings calculations
Developed RAG AI chatbot processing 50,000+ data points from 300+ courses with 95% accuracy. Reduced search time by 60% for 1000+ daily queries using Anthropic API and vector database.
- 95% accuracy with 50,000+ data points
- 60% faster search time reduction
- 1000+ daily queries processed
Built AI voice agent using VAPI API to bypass automated menus and connect with human representatives. Improved voice agent response latency by 70% implementing Groq's inference engine.
- Improved voice agent response latency by 70%
- Implemented Groq's inference engine
- Automated menu navigation system