senior AI Engineer

<h1 dir="auto">Senior AI Engineer – AI Vertical</h1><p dir="auto"><strong>Team:</strong> Engineering – AI Vertical<br><strong>Experience:</strong> 5+ years, including 2+ years building and shipping LLM systems in production<br><strong>Location:</strong> Mumbai – Andheri East<br><strong>Reports to:</strong> Head of Engineering<br><strong>Employment Type:</strong> Full-time</p><h2 dir="auto">About CashFlo</h2><p dir="auto">CashFlo is building the next generation of AI-powered Finance &amp; Accounting solutions for businesses.</p><p dir="auto">India sees over <strong>$11 trillion in B2B money flows every year</strong>, yet businesses typically wait around 70 days to get paid. CashFlo is transforming this ecosystem through technology across <strong>Payments, Lending, SaaS and AI</strong>.</p><p dir="auto">Our platform processes <strong>₹20,000+ Cr of invoices every month</strong>, serving <strong>300,000+ MSMEs and 1,200+ corporates</strong>, with 30+ lenders integrated for credit.</p><p dir="auto">CashFlo is now building a dedicated <strong>AI vertical for the global market</strong>, and we're looking for a Senior AI Engineer to join our core engineering team.</p><h2 dir="auto">About the Role</h2><p dir="auto">We are building AI agents that can operate core <strong>CFO workflows</strong> end-to-end — including accounts payable, month-end close, reconciliation, record-to-report, vendor management and compliance workflows.</p><p dir="auto">These agents don't just recommend actions. They <strong>execute finance processes</strong> by interpreting documents, applying accounting and control logic, performing transactions in enterprise systems such as SAP, and escalating genuine exceptions for human review.</p><p dir="auto">As a Senior AI Engineer, you will own systems <strong>end-to-end</strong>, from agent architecture and orchestration to production deployment, evaluation, reliability and scale.</p><h2 dir="auto">Key Responsibilities</h2><h3 dir="auto">1. Build Production AI Agents</h3><ul><li>Design and build autonomous agents for finance workflows such as invoice-to-pay, three-way matching, reconciliation, GL close and exception handling.</li><li>Build tool and action layers over ERP systems and enterprise applications.</li><li>Design approval workflows, human-in-the-loop checkpoints and segregation-of-duty controls.</li><li>Build robust error recovery and ensure complete auditability of agent actions.</li><li>Integrate with SAP APIs, ERP interfaces, documents and file-based systems.</li></ul><h3 dir="auto">2. Agent Memory &amp; Knowledge Systems</h3><ul><li>Design memory and context strategies for long-running agents.</li><li>Work on context windowing, summarization, compaction, retention and forgetting policies.</li><li>Build finance knowledge graphs covering vendors, invoices, POs, GRNs, GL accounts, contracts, approvals and provenance.</li><li>Work with graph databases such as <strong>Neo4j</strong> alongside vector search.</li></ul><h3 dir="auto">3. Evaluation &amp; Model Strategy</h3><ul><li>Build and improve evaluation frameworks for production AI systems.</li><li>Create golden datasets, workflow-level metrics and regression gates.</li><li>Develop automated evaluation and judging systems calibrated against human review.</li><li>Monitor model performance and detect drift on live workloads.</li><li>Work across <strong>OpenAI, Anthropic, Google and open-weight models</strong>.</li><li>Design model routing based on quality, latency and cost, including cascading and provider failover.</li></ul><h3 dir="auto">4. Scale &amp; Engineering</h3><ul><li>Own the latency, throughput and unit economics of AI workflows at scale.</li><li>Monitor and improve <strong>p50/p95/p99 latency</strong> and cost-per-workflow.</li><li>Build and maintain APIs using <strong>FastAPI and REST</strong>.</li><li>Mentor engineers working on AI systems and contribute to engineering best practices.</li></ul><h2 dir="auto">Required Experience</h2><ul><li><strong>5+ years of software/ML engineering experience</strong>, including <strong>2+ years shipping LLM systems in production</strong>.</li><li>Proven experience building and deploying <strong>multi-step AI agents</strong> that perform real actions in production systems.</li><li>Strong understanding of agent failure modes, reliability and control mechanisms.</li><li>Experience designing memory and context strategies for long-running agents.</li><li>Hands-on experience building an <strong>LLM evaluation framework</strong>, including datasets, metrics and automated judging.</li><li>Experience working with <strong>at least two LLM model providers or model families</strong>.</li><li>Experience migrating or operating live LLM workloads in production.</li><li>Experience operating AI/ML systems at significant scale with measurable production outcomes.</li></ul><h2 dir="auto">Technical Qualifications</h2><ul><li>Bachelor's/Master's degree in <strong>Computer Science, Computer Engineering or a related field</strong>.</li><li>Advanced proficiency in <strong>Python</strong>.</li><li>Strong experience with modern AI/ML development practices.</li><li>Hands-on experience with <strong>LangGraph, LangChain, OpenAI Agents SDK</strong>, or custom agent orchestration.</li><li>Working knowledge of <strong>graph databases</strong> such as Neo4j, Neptune or ArangoDB.</li><li>Understanding of graph data modelling, schema design and Cypher/Gremlin.</li><li>Strong API and data engineering fundamentals.</li><li>Experience with <strong>FastAPI, PostgreSQL and vector databases</strong>.</li><li>Experience deploying cloud-native applications on <strong>AWS, GCP or Azure</strong>.</li></ul><h2 dir="auto">Preferred Experience</h2><ul><li>Finance, accounting, audit or <strong>ERP/SAP</strong> domain exposure.</li><li>Experience with computer-use or browser automation agents.</li><li>Knowledge of <strong>GraphRAG and ontology design</strong>.</li><li>Experience with document AI/OCR.</li><li>Fine-tuning or serving open-weight models.</li><li>LLM observability and tracing using tools such as LangSmith, Langfuse or OpenTelemetry.</li><li>Understanding of enterprise AI security, governance and compliance.</li><li>Experience using AI coding agents as part of everyday software development.</li></ul><h2 dir="auto">Our Tech Stack</h2><p dir="auto"><strong>Python | LangGraph | LangChain | OpenAI | Anthropic | Google AI | FastAPI | PostgreSQL | pgvector | Neo4j | Elasticsearch | Java/Spring Boot | React | GCP | AWS | Docker</strong></p><p dir="auto">Familiarity with our exact stack is helpful but <strong>not mandatory</strong> if you have strong experience with equivalent technologies.</p><h2 dir="auto">What You'll Work On</h2><p dir="auto">You will directly contribute to building AI systems that:</p><ul><li>Execute real finance workflows instead of simply generating recommendations.</li><li>Interact with enterprise systems such as SAP.</li><li>Process large volumes of financial documents and transactions.</li><li>Make finance operations more automated and efficient.</li><li>Operate with strong controls, auditability and human oversight.</li><li>Serve enterprise customers at scale.</li></ul><h2 dir="auto">Why Join CashFlo?</h2><ul><li><strong>Build from the ground up:</strong> Be part of a new AI vertical targeting the global market.</li><li><strong>High-impact work:</strong> Your systems will operate on real enterprise finance workflows at scale.</li><li><strong>Strong engineering culture:</strong> Work alongside experienced engineers and leadership from top consulting, technology and finance backgrounds.</li><li><strong>Ownership:</strong> Own problems end-to-end rather than individual components.</li><li><strong>Growth:</strong> Opportunity to work across AI agents, LLMs, enterprise systems, evaluation and production infrastructure.</li><li><strong>Compensation &amp; Equity:</strong> Competitive compensation, incentive structures and meaningful equity in an early-stage company.</li></ul>