AI Product Manager

Why this role exists

Were scaling our AI product suite - invoice & contract and terms extraction, anomaly and overpayment detection, and analyst workflows for both internal and client users. We need a product leader who can turn ambiguous customer pain into shipped, secure, and measurable AI features that enterprise customers love and finance leaders trust.

What youll own

Strategy & Outcomes

• Define and own the AI product strategy and measurable outcomes for recovery, accuracy, and time-to-value across enterprise accounts.
• Translate market and customer signals (Procurement, AP, Audit, Infosec, Legal) into a prioritized roadmap with clear trade-offs.

Product Discovery

• Run deep discovery with Fortune 1000 stakeholders and Discover Dollar analysts to map jobs-to-be-done, decision points, and data realities (SAP/Oracle/Coupa/Ariba exports, data lake, data warehouse extracts, contracts, invoices, emails, credit notes).
• Validate problems and solutions with prototypes (AI tools like Lovable, Replit, Figma, Low-code tools etc) and pilot metrics before committing engineering.

Product Delivery

• Lead cross-functional squads (ML/DS, Backend, Frontend, Design, Data Platform, Security) to ship high-quality increments on a predictable cadence.
• Write crisp PRDs, acceptance criteria, and instrument the product for analytics/telemetry (events, funnels, precision/recall dashboards).
• Drive integrations and partner workflows (e.g., SAP S/4HANA, Oracle, Coupa/Ariba, RMS) with clear APIs and data contracts.

AI/ML Product Craft

• Shape AI features such as document extraction (OCR + NLP), contract terms understanding, Complex Agentic AI workflows, LLM/RAG assistants for analysts, anomaly/risk scoring, and human-in-the-loop review.
• Partner with DS/ML on evaluation frameworks (offline/online), data labeling strategy, prompt & model versioning, and cost/performance tuning (latency, tokens, GPU/CPU budgets).
• Establish monitoring/guardrails: drift, bias, hallucination containment, red-teaming, feedback loops.

Governance, Privacy & Security

• Embed privacy-by-design and enterprise controls (role-based access, audit logs, retention, encryption).
• Coordinate with customers InfoSec/Legal on DPAs, SOC 2/ISO 27001 alignment, data residency, and model risk documentation (model cards, DPIAs where needed).

Customer & GTM

• Own product pilots, success criteria, and value realization (business cases, ROI).
• Enable Sales/CS with narratives, demos, pricing/packaging inputs, and win/loss insights.
• Collect structured feedback and convert it into backlog and experiments.

What success looks like

30 days: Understand domain and data flows; publish product health baseline; ship a small improvement (e.g., analyst UX, cost reduction).
60 days: Deliver a validated AI feature MVP (e.g., improved clause extraction or anomaly ranking) to a pilot customer with a signed success plan.
90 days: Hit agreed accuracy + adoption targets; publish a 2-quarter roadmap tied to revenue and recovery dollars; formalize AI evaluation & monitoring dashboards.

Key KPIs/metrics youll move

• Business impact: Verified recovery dollars, time-to-detect, time-to-close.
• Model quality: Precision/recall/F1 by use case; false positive rate; drift indicators.
• Efficiency: Cost per decision (tokens/compute), latency, analyst throughput.
• Adoption: Active users, task completion rate, retention, NPS/CSAT.
• Reliability & risk: Incident rate, hallucination guardrail triggers, SLA/SLO adherence.

Must-have qualifications

• 5+ years in product management with at least 1 year of shipping AI/ML-powered features in production (LLMs/RAG/NLP/vision/ML detection).
• Track record in B2B enterprise SaaS with complex data and security needs; comfortable working with Procurement/AP/Finance stakeholders.
• Strong discovery and storytelling skills: can turn messy datasets and ops realities into a clear product narrative and roadmap.
• Technical depth to partner with DS/ML/Engineering understands evaluation metrics, A/B testing, data pipelines, API design, and basics of SQL/Python.
• Analytics-first: habit of instrumenting funnels, defining guardrails, and making decisions from dashboards and user studies.
• Excellent written communication (PRDs, RFCs, customer docs) and stakeholder management across Sales, CS, Legal, and Infosec.
• Bachelors in Engineering/CS/Math/Statistics or equivalent experience.

Nice-to-have

• Domain exposure to Source-to-Pay, AP automation, recovery audit, contract analytics, or financial controls.
• Experience with SAP/Oracle/Workday/Coupa/Ariba data models and enterprise integrations.
• Familiarity with MLOps (MLflow/W&B), vector databases (FAISS/Pinecone), and eval frameworks for LLMs.
• Understanding of model risk management, secure AI patterns, and enterprise compliance expectations.
• Prior startup or zero-to-one product experience.

Our current tech context (helpful, not required)

• AI Stack: LLMs (commercial & open-source Open AI, Llama, Claude), RAG pipelines, OCR + NLP for documents,
• Data stack: cloud object storage, relational DBs, event pipelines, analytics stack like Databricks, Elastic Search,
• Cloud Stack: Primarily Azure but build cloud agnostic product to make it deployable in AWS, GCP etc.
• Dev Stack: Python, Javascript, React
• Product tooling: Clickup (Jira equivalent), Postman, Figma

How we work

• Bangalore-based hybrid team with collaboration across India and North America
• Pragmatic agile: weekly planning, monthly business reviews, quarterly product reviews tied to revenue and recovery outcomes.

What youll get to do

• Ship AI that recovers real money for global enterprises. (Already saved over $1B+ with current product, need to 10X the impact)
• Build analyst-in-the-loop systems that blend human judgment with AI scale.
• Influence pricing/packaging and help define the category for AI-assisted recovery audit.

Compensation & benefits

Competitive with Bangalore market for Product Manager talent, plus benefits and performance incentives. (Well share details during the first call.)