Designation: Agentic AI Engineer – ERP Automation
Department: Technology / Product Engineering
Experience: 3–7 Years (flexible based on expertise)
Location: Bangalore
1. Role Overview
We are looking for an experienced Agentic AI Specialist who can transform an existing ERP application into an AI-enabled, intelligent business operations platform.
The individual will be responsible for understanding existing application workflows, identifying opportunities for automation, and developing AI-powered agents that can independently perform tasks currently handled by human users.
The primary focus will be on automating accounting and financial operations by enabling AI agents to read, understand, interpret, validate, and process financial documents and automatically perform the necessary accounting entries within the existing ERP system.
The objective is to build an ERP platform where approximately 70% of routine operational and accounting activities can be automated through AI, with human intervention reserved for exceptions, approvals, and complex decisions.
2. Key Responsibilities
A. Agentic AI Development
- Design, develop, and deploy autonomous AI agents capable of executing multi-step business workflows.
- Build intelligent agents that can understand instructions, make contextual decisions, execute tasks, and verify outcomes.
- Implement multi-agent architectures where specialized agents collaborate to complete accounting and operational processes.
- Develop AI agents that can interact with existing ERP modules through APIs and application interfaces.
- Enable agents to learn from historical transactions, accounting patterns, business rules, and user corrections.
- Implement human-in-the-loop mechanisms for approvals and exception handling.
B. AI-Powered Accounting Automation
Develop an intelligent accounting engine capable of:
- Reading and interpreting bills, invoices, receipts, bank statements, and other financial documents.
- Extracting relevant information, including vendor details, invoice dates, amounts, tax information, line items, and payment terms.
- Identifying the nature of each transaction and understanding its accounting implications.
- Automatically identifying the appropriate General Ledger (GL) accounts and chart of accounts.
- Determining debit and credit entries based on the nature of the transaction.
- Identifying applicable tax treatments and accounting classifications.
- Creating journal entries and posting transactions into the existing ERP.
- Matching bills with purchase orders, payments, and vendor records.
- Detecting duplicate invoices, incorrect amounts, missing information, and unusual transactions.
- Reconciling transactions and identifying discrepancies.
- Learning from accountant corrections and previously approved accounting entries to improve accuracy.
C. Existing ERP Application Automation
- Study the existing ERP architecture, application workflows, database structure, and APIs.
- Identify repetitive manual activities across accounting, finance, operations, and administration.
- Convert existing manual workflows into AI-driven automated workflows.
- Integrate AI capabilities into existing ERP modules without disrupting current functionality.
- Build intelligent search, natural-language commands, automated reporting, and workflow orchestration.
- Enable users to interact with the ERP using conversational AI.
- Automate data entry, transaction processing, reconciliation, reporting, and routine follow-ups.
D. AI Integration and Engineering
- Integrate Large Language Models (LLMs) such as OpenAI, Anthropic, or equivalent models.
- Work with AI frameworks such as LangChain, LangGraph, LlamaIndex, or equivalent technologies.
- Build RAG-based systems to enable AI to retrieve information from ERP databases, accounting policies, historical transactions, and business documents.
- Develop document intelligence capabilities using OCR, document parsing, and vision models.
- Build secure API integrations and backend automation.
- Optimize AI model performance, response times, and inference costs.
- Implement logging, monitoring, traceability, and error recovery mechanisms.
E. Testing, Accuracy and Governance
- Establish accuracy benchmarks for document extraction and accounting classifications.
- Ensure that every AI-generated accounting entry is traceable to its source document.
- Implement validation rules to prevent incorrect or duplicate postings.
- Establish approval workflows for high-value or unusual transactions.
- Maintain audit trails of AI decisions, modifications, and postings.
- Ensure data privacy, access controls, and financial data security.
3. Technical Skills Required
Must-have skills:
- Strong experience in Python and backend development.
- Hands-on experience building AI agents and agentic workflows.
- Practical experience with LLMs and prompt engineering.
- Experience with OpenAI APIs, Anthropic, or similar AI platforms.
- Experience with LangGraph, LangChain, LlamaIndex, or similar frameworks.
- Knowledge of REST APIs, webhooks, and application integrations.
- Experience with OCR, document intelligence, and structured data extraction.
- Knowledge of SQL and relational databases.
- Experience building production-ready AI applications.
Good-to-have skills:
- Understanding of accounting workflows, double-entry bookkeeping, and General Ledger.
- Experience integrating AI with ERP, accounting, or financial applications.
- Experience with QuickBooks, Xero, NetSuite, Zoho Books, or similar systems.
- Knowledge of vector databases and RAG architectures.
- Experience with AWS, Azure, or Google Cloud.
- Experience with workflow automation tools and event-driven architectures.
Pay: ₹1,500,000.00 - ₹1,800,000.00 per year
Benefits:
- Health insurance
- Provident Fund
Work Location: In person
