AI Agent Engineer - Governed Agentic Systems

The Enterprise; Hyderabad; On-site; INR 800,000 - 1,100,000; Full time

Job summary

The Enterprise is seeking an experienced AI Agent Engineer in Hyderabad, India for a full-time hybrid role. You will design and develop multi-agent systems for industries like finance and healthcare, ensuring predictable agent behavior within production limits.

The ideal candidate has 3–6 years of ML or software engineering experience, strong Python skills, and familiarity with frameworks such as LangGraph or AutoGen. This hands-on role emphasizes designing agent workflows and integrating them with customer processes.

Qualifications

• 3–6 years of software or ML engineering experience.
• Hands-on experience with agentic frameworks and Python skills.
• Strong communication skills with non-technical stakeholders.

Responsibilities

• Design and implement multi-agent architectures.
• Define agent boundaries and escalation paths.
• Build and maintain memory systems for agent tasks.
• Instrument agent traces and action logs.
• Integrate agents with customer engineering teams.

Skills

Software Engineering

Machine Learning

Python

Agentic Frameworks

Communication Skills

Tools

LangGraph

AutoGen

CrewAI

Job description

AI Agent Engineer - Governed Agentic Systems

Hyderabad, India – Hybrid, Full‑time, Senior level

About the role

Enterprises need agents that behave predictably in production. As an AI Agent Engineer at ArqAI you will design and ship multi‑agent systems for customers in financial services, insurance, and healthcare. All actions must be traceable, decisions bounded, and the system safe to run. You will own the orchestration layer: how agents communicate, when they escalate, how they fail gracefully, and how the customer’s team audits what happened after the fact. This is a hands‑on engineering role, not a research or consulting position.

What you'll do

• Design and implement multi‑agent architectures, orchestrators, sub‑agents, and tool‑calling layers using frameworks such as LangGraph, AutoGen, or CrewAI aligned with real customer workflows.
• Define agent boundaries, escalation paths, human‑in‑the‑loop checkpoints, and fallback logic so the system degrades gracefully when confidence is low.
• Build and maintain tool registries, memory systems (short‑term, episodic, and long‑term), and context‑management strategies for agents operating over extended tasks.
• Instrument agent traces, action logs, and decision rationale so customer teams can audit, replay, and improve agent behaviour post‑deployment.
• Work directly with customer engineering teams to integrate agents into their existing identity, data, and approval workflows.

What we're looking for

• 3–6 years of software or ML engineering experience, with at least 1 year building and shipping agentic or autonomous LLM systems in production.
• Hands‑on experience with LangGraph, AutoGen, CrewAI, or a comparable agentic framework and an opinion on when each earns its complexity.
• Strong Python skills with experience in async patterns, tool/function‑calling APIs, and structured output schemas.
• Working knowledge of agent observability: trace capture, span logging, prompt/response auditing, and cost tracking per agent run.
• Clear communication skills, able to explain an agent’s decision path to non‑technical stakeholders without hand‑waving.

Nice to have

• Experience in regulated industries where agent actions have compliance or legal implications.
• Familiarity with OpenAI Assistants API, Anthropic tool use, or Gemini function calling at production scale.
• Background in process automation, RPA, or workflow orchestration prior to the LLM era.