AI Agent Developer

MacroCoderz; Pakistan; Remote; PKR 2,000,000 - 4,200,000; Full time

Job summary

MacroCoderz is hiring a Remote AI Agent Developer to design and build LLM agents that call tools and APIs to complete multi-step business tasks with human approval.

In this role you will define the tools an agent can use, build guardrails, and implement human-in-the-loop workflows that keep humans in control while agents operate autonomously when appropriate.

You will create evaluation scenarios, trace agent runs in production, and optimize cost and latency per task across client projects.

Qualifications

• 2+ years of software engineering in Python or TypeScript.
• Has built at least one agent that calls tools or APIs - professionally or as a serious project.
• Understands function calling, structured outputs and how agents fail.
• Comfortable working with REST APIs, webhooks and authentication flows.

Responsibilities

• Design and build tool-using LLM agents and multi-agent workflows.
• Write tool definitions and integrations against client APIs and SaaS platforms.
• Build guardrails, permission checks and human-in-the-loop approval steps.
• Create evaluation scenarios that test agents on realistic, messy inputs.
• Trace and debug agent runs in production, and reduce cost and latency per task.

Skills

Python

TypeScript

Agent design

REST APIs

Tools

OpenAI Agents SDK

LangGraph

API integrations

Job description

Remote AI Agent Developer role at MacroCoderz: build LLM agents that use tools, call APIs and complete multi-step business tasks with human approval.

AI agents are the fastest-growing part of our work. Clients want systems that do more than answer questions: agents that triage support tickets, update CRMs, draft documents for review, and hand off to a person when they are unsure.

As an AI Agent Developer you design the tools an agent can use, the guardrails around them, and the approval steps that keep a human in control - then prove the agent works with evaluation runs before it reaches users.

What you will do

• Design and build tool-using LLM agents and multi-agent workflows.
• Write tool definitions and integrations against client APIs and SaaS platforms.
• Build guardrails, permission checks and human-in-the-loop approval steps.
• Create evaluation scenarios that test agents on realistic, messy inputs.
• Trace and debug agent runs in production, and reduce cost and latency per task.

What we are looking for

• 2+ years of software engineering in Python or TypeScript.
• Has built at least one agent that calls tools or APIs - professionally or as a serious project.
• Understands function calling, structured outputs and how agents fail.
• Comfortable working with REST APIs, webhooks and authentication flows.

Nice to have

• Experience with LangGraph, the OpenAI Agents SDK, Claude tool use or a similar agent framework.
• Familiarity with the Model Context Protocol (MCP).
• Background in workflow automation or backend integrations.

Location

fully remote, open to applicants worldwide.