About the role

Team: Engineering – AI Vertical Experience: 5+ years, including 2+ years building and shipping LLM systems in production Location: Mumbai – Andheri East Reports to: Head of Engineering Employment Type: Full-time
About The Role We are building AI agents that can operate core CFO workflows end-to-end — including accounts payable, month-end close, reconciliation, record-to-report, vendor management and compliance workflows. These agents don't just recommend actions. They execute finance processes by interpreting documents, applying accounting and control logic, performing transactions in enterprise systems such as SAP, and escalating genuine exceptions for human review. As a Senior AI Engineer, you will own systems end-to-end from agent architecture and orchestration to production deployment, evaluation, reliability and scale.
Key Responsibilities

Build Production AI Agents
Design and build autonomous agents for finance workflows such as invoice-to-pay, three-way matching, reconciliation, GL close and exception handling.
Build tool and action layers over ERP systems and enterprise applications.
Design approval workflows, human-in-the-loop checkpoints and segregation-of-duty controls.
Build robust error recovery and ensure complete auditability of agent actions.
Integrate with SAP APIs, ERP interfaces, documents and file-based systems.
Agent Memory & Knowledge Systems
Design memory and context strategies for long-running agents.
Work on context windowing, summarization, compaction, retention and forgetting policies.
Build finance knowledge graphs covering vendors, invoices, POs, GRNs, GL accounts, contracts, approvals and provenance.
Work with graph databases such as Neo4j alongside vector search.
Evaluation & Model Strategy
Build and improve evaluation frameworks for production AI systems,
Create golden datasets, workflow-level metrics and regression gates.
Develop automated evaluation and judging systems calibrated against human review.
Monitor model performance and detect drift on live workloads.
Work across OpenAI, Anthropic, Google and open-weight models.
Design model routing based on quality, latency and cost, including cascading and provider failover.
Scale & Engineering
Own the latency, throughput and unit economics of AI workflows at scale.
Monitor and improve p50/p95/p99 latency and cost-per-workflow.
Build and maintain APIs using FastAPI and REST.
Mentor engineers working on AI systems and contribute to engineering best practices.

Required Experience

5+ years of software/ML engineering experience, including 2+ years shipping LLM systems in production.
Proven experience building and deploying multi-step AI agents that perform real actions in production systems.
Strong understanding of agent failure modes, reliability and control mechanisms.
Experience designing memory and context strategies for long-running agents.
Hands-on experience building an LLM evaluation framework, including datasets, metrics and automated judging.
Experience working with at least two LLM model providers or model families.
Experience migrating or operating live LLM workloads in production.
Experience operating AI/ML systems at significant scale with measurable production outcomes.

Technical Qualifications

Bachelor's/Master's degree in Computer Science, Computer Engineering or a related field.
Advanced proficiency in Python.
Strong experience with modern AI/ML development practices.
Hands-on experience with LangGraph, LangChain, OpenAI Agents SDK, or custom agent orchestration.
Working knowledge of graph databases such as Neo4j, Neptune or ArangoDB.
Understanding of graph data modelling, schema design and Cypher/Gremlin.
Strong API and data engineering fundamentals.
Experience with FastAPI, PostgreSQL and vector databases.
Experience deploying cloud-native applications on AWS, GCP or Azure.

Preferred Experience

Finance, accounting, audit or ERP/SAP domain exposure.
Experience with computer-use or browser automation agents.
Knowledge of GraphRAG and ontology design.
Experience with document AI/OCR.
Fine-tuning or serving open-weight models.
LLM observability and tracing using tools such as LangSmith, Langfuse or OpenTelemetry.
Understanding of enterprise AI security, governance and compliance.
Experience using AI coding agents as part of everyday software development.

Execute real finance workflows instead of simply generating recommendations
Interact with enterprise systems such as SAP.
Process large volumes of financial documents and transactions.
Make finance operations more automated and efficient.
Operate with strong controls, auditability and human oversight.
Serve enterprise customers at scale.