Build Secure Agentic AI Integration Layer for Enterprise Risk Management Platform

Job ID: 40227061

Budget: ₹1,500 – ₹12,500 INR

We are looking for an experienced AI/Backend Engineer (or small team) to help design and implement a foundational setup for an enterprise “Agentic AI” assistant.The goal is to enable a prompt-based interface where business users can ask questions and trigger actions across internal risk management and document systems through a governed API layer.

This is NOT a chatbot UI project.This project focuses on backend architecture, AI orchestration, and secure enterprise integration.

Target application includes IBM OpenPages SaaS (GRC) accessed via REST APIs.

The first phase will implement a proof-of-concept that allows an AI agent to retrieve and update structured records using approved APIs via a secure middleware service.

Scope of Work (Phase-1: Foundational Setup)

Build a secure middleware API service (AI Gateway)

Integrate an enterprise LLM (Azure OpenAI or equivalent hosted LLM)

Implement tool/function calling (agent actions)

Connect to an existing enterprise system via REST APIs (query + update operations)

Implement structured response schema (business entities like risks, issues, controls, tasks)

Implement logging, traceability, and error handling

Implement role-aware request handling (no hardcoded credentials)

Ensure deployment architecture supports IP allow-listing

Provide clear documentation and deployment instructions

No UI required (basic Postman/test interface sufficient)

Expected Deliverables

Backend AI orchestration service (production-style codebase)

Tool schemas and function calling implementation

API mapping layer (business model → system APIs)

Sample prompts and evaluation tests

Security design notes

Deployment guide (cloud-ready)

Architecture diagram

Required Technical Skills

Core Backend & APIs

Python (FastAPI / Flask) OR Node.js backend development

REST API integration and schema mapping

OAuth / token-based authentication handling

Secure error handling and structured logging

AI / LLM

OpenAI / Azure OpenAI APIs

Function calling / tool calling agents

Prompt engineering for structured outputs

Retrieval-augmented workflows (optional but preferred)

Cloud & Infrastructure

Azure (preferred) or AWS/GCP deployment

Static outbound IP / networking configuration

Containerization (Docker)

Environment configuration management

Architecture Knowledge

Designing middleware integration layers

Converting complex APIs into business semantic APIs

Handling long-running workflows

Designing safe update operations (write vs read actions)

Nice to Have

Experience in enterprise systems integration

Experience with agent frameworks (LangChain, LangGraph, Semantic Kernel, or similar)

Knowledge of approval workflows / audit logging

Experience deploying internal copilots or assistants

Responsibilities

Design the AI orchestration pattern (not just code implementation)

Implement secure tool-based action execution

Prevent unsafe or unintended write operations

Provide extensible architecture for future integrations

Document how new tools/actions can be added later

Ensure solution is maintainable by internal developers

Engagement TypeShort initial phase (2–4 weeks) with potential long-term extension for full platform buildout.

When Applying, Please Include

Relevant projects involving LLM integrations or AI agents

Example of tool/function calling implementation (code or architecture)

Cloud deployment experience

Preferred tech stack and why

Estimated timeline for Phase-1

We are specifically looking for someone who understands enterprise-grade AI integration, not only prompt engineering or UI chatbot development