Hi, I'm Harish Govindaswamy Krishnamurthy 👋
AI Automation Architect | Senior SDET | Agentic Systems Engineer
I specialize in building the bridge between AI-native engineering and enterprise-grade reliability. My focus is on shifting quality "ultra-left" by using Agentic LLM workflows, MCP servers, and high-concurrency automation frameworks to evaluate and deploy safe AI systems.
🚀 Featured Repositories
🤖 agentic-eval-orchestrator
AI & Agentic Systems | Python, LangGraph, MCP, Ollama
- An end-to-end Agentic Evaluation System designed to benchmark LLM outputs and identify model failure modes.
- Key Features: Uses LangGraph for multi-agent orchestration; integrates MCP servers for secure data context; leverages Ollama for local, privacy-focused inference.
- Impact: Automates Synthetic Data Generation and scales model stress testing for production-grade AI deployments.
🎭 playwright-enterprise-framework
Test Engineering | TypeScript, Playwright, GitHub Actions
- A production-grade, high-concurrency automation framework designed for "ultra-left" quality shifting and performance profiling.
- Key Features: Advanced Playwright tracing, custom API mocking, and a parallel execution engine optimized for CI/CD.
- Impact: Scaled to support 2,000+ daily executions, reducing regression feedback loops from 12 hours to 90 minutes.
⛓️ solana-lending-protocol
Web3 & Blockchain Engineering | Rust, Anchor Framework, Solana CLI
- A decentralized lending and borrowing protocol built on the Solana blockchain.
- Key Features: Implemented cross-program invocations (CPIs), interest rate logic, and collateralization checks.
- Testing: Comprehensive Solana Test Suite using TypeScript/Mocha to validate transaction atomicity, liquidations, and instruction security.
🛠️ Technical Toolkit
| Category | Tools & Technologies |
|---|---|
| AI & Agents | LangGraph, CrewAI, MCP, Ollama, OpenAI API, Azure AI Foundry |
| Languages | Python (Production), TypeScript, Rust (Anchor/Solana), SQL, C# |
| Automation | Playwright, PyTest, API Mocking (WireMock/Playwright Mocks) |
| Evaluation | Synthetic Data Gen, RAG Evaluation, Model Failure Mode Analysis |
| DevOps | GitHub Actions, Azure DevOps (YAML), Docker, Kubernetes |
📈 Professional Highlights
- Architected a multi-agent pipeline using Python/LangGraph to automate 500+ complex test suites.
- Developed secure lending smart contracts on Solana, bridging Fintech logic with Web3 engineering.
- Scaled internal AI prototyping using Ollama, ensuring data privacy and slashing API overhead.
- Reduced regression cycle time by 87% through high-concurrency Playwright architectures.
📫 Connect with Me
- LinkedIn: [linkedin.com/in/your-profile]
- Email: blkcdev@gmail.com
- Location: Santa Clara, CA
“Building the systems that ensure AI is as reliable as the hardware it runs on.”