Looking for a job

Mihir Saini

@mihirsaini

Message

I build production agentic AI systems and scalable full-stack platforms.

India

Message

What I'm looking for

I'm looking to build production-grade agentic AI systems and scalable full-stack platforms, taking products from proof of concept to reliable, high-impact deployment.

At Eagleview, I lead delivery of Horizon, a production multi-agent system built with Strands workflows, MCP, AG-UI streaming, Bedrock guardrails, DuckDB analytics, and execution tracing.

I re-architected Horizon property search into MCP-backed tools, reducing prompt footprint by 68%, token consumption by 56%, and average model cost per workflow by 49%. I also built a RAG-based Jira triage agent that reduced mean time to assignment by 40%.

I scaled Eagleview One’s micro-frontend platform from proof of concept to production, enabling five teams and 30+ engineers to independently maintain and deploy more than five applications. I built EV Map SDK and led the Property Viewer MVP, helping successful order loads reach 99%.

I also build end-to-end products, including Reeler, an outbound webhook platform, VidScribe, a cross-platform YouTube note-taking system, and DeepScene, a deepfake-video classification application. I won Eagleview’s company-wide hackathon for NeuraMap, an agentic mapping solution incorporated into Horizon.

Experience

Work history, roles, and key accomplishments

Current

Software Engineer II

Current

Eagleview

Nov 2025 - Present (10 months)

Lead technical delivery across an 8+ engineer team building Eagleview Horizon, a production multi-agent system. Re-architected property search to reduce prompt footprint by 68% and token consumption by 56%.

MCP AG UI Bedrock DuckDB Micro Frontends RAG Jira TypeScript NodeJS AWS

Software Engineer I

Eagleview

Jul 2024 - Nov 2025 (1 year 4 months)

Designed foundational micro-frontend architecture and built EV Map SDK as a modular monorepo. Led delivery of Property Viewer MVP, increasing successful order loads to 99%.