Siamak S.

I build software in an AI-native way—and I build the AI systems that make it possible. I'm an AI-native software engineer and former startup CTO/co-founder. I work across modern AI coding harnesses and agent frameworks to design, build, and ship complex software workflows—not just demos or isolated chatbots. My work spans three connected areas: **AI agents and complex workflows** I build tool-using agents and multi-step workflows that connect models to APIs, data, and business systems. I focus on orchestration, state, reliability, observability, human approval, and the engineering needed to make workflows useful in real environments. **RAG and knowledge systems** I build retrieval-augmented applications that connect LLMs to documents, databases, and APIs. This includes ingestion, retrieval, vector search, grounded responses, and practical evaluation. **Real-time voice and conversational systems** In the Omvi project, I worked on complex, real-time voice agents. I’m interested in the engineering challenges behind responsive conversational experiences and agents that can operate reliably in live interactions. **AI-native engineering and team workflows** I experiment with different AI coding harnesses and agent-assisted development approaches. My current focus is developing better systems for managing AI-native software teams: how people and agents coordinate, how work is specified and reviewed, and how teams maintain quality while keeping pace with a rapidly changing industry. I also bring hands-on product and backend experience: - TypeScript, Node.js, Python, and FastAPI - React and Next.js - PostgreSQL, vector search, and API integrations - Agent frameworks, RAG architectures, and MCP - AI-assisted software development and workflow automation As a former CTO and product builder, I can work from an ambiguous problem through architecture, implementation, and delivery. I care about clear scope, maintainable systems, honest trade-offs, and software that solves a real operational problem. **Good projects to start with** - Building or extending an AI agent or multi-step workflow - Creating a RAG assistant over company knowledge - Developing a real-time voice agent - Integrating AI into an existing product or backend - Designing an AI-native development workflow for an engineering team If you have a technically challenging AI product or workflow in mind, tell me what you’re trying to achieve and what you’ve built so far. I’ll help identify a practical engineering path.