AI Agent Developer

Agilysis; Hyderabad, Pune District, Bengaluru; On-site; INR 900,000 - 1,900,000; Full time

Job summary

Agilysis is seeking a hands-on AI Agent Developer to build, integrate and support AI agents that automate business workflows. You will work with the AI Architect and Product Owner to translate requirements into reliable, secure solutions from prototype to production and continuous improvement.

You will implement agent logic, prompts, tool calls and orchestration, while ensuring test coverage, security, and maintainable code through delivery pipelines.

Qualifications

• Hands-on software development in Python or TypeScript with testing and CI/CD.
• Practical experience building tool-enabled AI agents and prompts.
• Ability to translate workflows into executable logic with deterministic rules.

Responsibilities

• Implement and orchestrate AI agents and workflows with strong custodianship.
• Develop connectors to enterprise applications and APIs with secure practices.
• Create tests and evaluation scenarios for end-to-end agent behavior.
• Ensure security, least-privilege access and robust logging in production.

Skills

Python

TypeScript

LLMs

APIs

CI/CD

Automated testing

Tools

REST APIs

OAuth

SQL/NoSQL

Embeddings

Job description

Job Summary

We are looking for a hands‑on AI Agent Developer to build, integrate and support AI agents that execute defined business workflows. Working closely with the AI Architect and Product Owner, this person will translate workflow requirements and approved architecture into reliable, secure and maintainable solutions, taking agents from prototype through testing, production deployment and continuous improvement.

Responsibilities

• Workflow implementation. Work with the Product Owner to understand business processes, user needs, exceptions and acceptance criteria. Translate these into agent logic, workflow steps and technical implementation tasks, aligned with the AI Architect's design.
• Agent development and orchestration. Build agents using prompts, tool calling and workflow orchestration. Implement state management, routing, validation, retries, timeouts and fallback paths, including human review and approval where required. Develop connectors and tools to interact with enterprise applications and APIs.
• RAG and enterprise knowledge. Implement document ingestion, chunking, embeddings, retrieval and grounded responses with source citations where needed. Apply source access permissions and support updates to keep agent knowledge current.
• Testing and evaluation. Create automated tests and representative evaluation scenarios covering normal workflows, exceptions and misuse. Measure response quality, task completion, tool execution accuracy, latency and cost against agreed acceptance criteria.
• Security and responsible use. Implement controls defined by the AI Architect and security teams, including least‑privilege access, sensitive‑data protection, prompt‑injection defenses, audit trails and human approval before consequential actions.
• Deployment and production support. Package and deploy agents through approved delivery pipelines. Implement logging, tracing and monitoring; investigate failures and support incident resolution, version management and rollback.
• Continuous improvement and documentation. Improve agent reliability, performance and cost based on evaluations and user feedback. Maintain code, tool contracts, configuration, technical documentation and support runbooks.

Required experience and skills

• Software development. Strong hands‑on development skills in Python or TypeScript, with experience building maintainable applications using version control, code reviews, automated testing and CI/CD.
• LLMs and agent development. Practical experience building tool‑enabled AI agents, including prompting, structured outputs, context management and orchestration. Understanding of model limitations, unreliable outputs and agent failure modes.
• Workflow engineering. Ability to translate business workflows into executable logic, including conditional paths, approval steps, exception handling and recovery. Ability to identify where deterministic rules should support agent behavior.
• APIs and integration. Experience with REST APIs, authentication methods such as OAuth, SQL/NoSQL databases and asynchronous processing. Ability to build reliable integrations with existing enterprise systems.
• Evaluation and troubleshooting. Ability to build test datasets, evaluate end‑to‑end agent behavior and diagnose issues across prompts, retrieval, tools and integrations.
• Collaboration and delivery. Ability to work closely with AI Architects, Product Owners and business stakeholders, communicate implementation constraints and deliver iteratively in a fast‑paced environment

Preferred experience

• Knowledge of HR processes and enterprise HR platforms such as Workday, including employee‑data access, workflow integration and approval boundaries.
• Experience with enterprise agent platforms, orchestration frameworks, cloud services and AI evaluation or observability tools.