AI Automation Engineer — Agentic Builds
Role Overview
This is a hands-on building role for someone who lives at the intersection of business-process engineering and applied AI: you will map how work flows today, redesign it around automation, and then build the agents and orchestration that make the current process run.
You will own builds end to end — from process discovery through to agents deployed on Azure (Container Apps and Function Apps) and integrated into the company's broader systems and orchestration layer. You are equally comfortable reasoning about a messy operational workflow and writing the code that automates it, using modern agent-building tooling such as Codex and Claude Code.
What You'll Do
Connect business process to automation. Translate operational workflows into AI-automation opportunities and identify where agents create the most leverage.
Process map, then reengineer. Build process maps of current-state workflows and use them to redesign and reengineer processes for automation — not just digitizing the old way of working.
Build agents from the ground up.
Architect and code agents from scratch using tools such as Codex and Claude Code, with clean, maintainable, testable implementations.
Ship agentic builds on Azure. Deploy and run agents in Azure Container Apps and Azure Function Apps, including packaging, configuration, secrets, scaling, and monitoring.
Design for the larger ecosystem. Build agents that fit the company's broader systems and orchestration layer — multi-agent coordination, tool and data integrations, state, and handoffs across the ecosystem.
Optimize cost and performance. Make deliberate model-selection and cost-optimization decisions, balancing capability, latency, and spend across model tiers and routing strategies.
Make it production-ready. Add evaluation, observability, error handling, and guardrails so builds are reliable, measurable, and safe to run against real work.
What You Bring (Required)
Demonstrated experience buil
📌 Automation AI Engineer (India)
🏢 Straive
📍 India