Job Description

Engineering Manager, AI/ML Innovation & Product Engineering

Job Title

Engineering Manager, AI/ML Innovation & Product Engineering

Function

BPS Innovation Lab Technology Build Team

Reporting To

Innovation & Growth Lab Head

Location

India Pune

Team Size

Leads the Labs tech team (AI/ML + full-stack, headcount grows with hiring)

Grade / Level

Manager

Employment Type

Full-Time

ROLE PURPOSE

The Engineering Manager, AI/ML Innovation & Product Engineering is the most critical technical hire in the Innovation Lab. This individual will design the architecture, build the initial versions, and establish the engineering standards for all AI-enabled and full-stack products developed by the Lab. Working closely with the Lab Head on commercial direction and leading the Labs tech team on execution, this role is the bridge between a validated business concept and a client-deployable product.

This role is ideal for a leader who is passionate about designing cutting-edge AI solutions, driving innovation, and delivering measurable business impact for global clients. The candidate should bring a builder’s mindset, hands-on expertise in AI/ML systems and Agentic AI, and the ability to translate complex client problems into scalable AI-powered solutions.

KEY RESPONSIBILITIES

1. Architecture & Technical Design

- Design the end-to-end technical architecture for all AI/ML and full-stack products built by the Lab, including agentic workflows, API integrations, service architecture, and inference infrastructure
- Lead solutioning discussions with global stakeholders to define AI-driven transformation roadmaps
- Define and enforce engineering standards, code review protocols, and documentation practices across the full tech team (AI/ML and full-stack)
- Evaluate and select the appropriate AI frameworks, cloud infrastructure, and third-party APIs (including LLM providers) for each product
- Ensure architecture decisions account for client data residency, security requirements, and third-party processing constraints from Day 1
- Own the IP integrity of all technical work — ensuring all code, models, and build artefacts are developed under ownership framework

2. AI Data Engineering & Pipelines

- Own data pipelines for multi-modal and unstructured inputs — receipts, PDFs, vendor statements, contracts, email — across all Lab products
- Build OCR/layout-parsing ingestion and structured-extraction pipelines from unstructured sources
- Design and maintain embedding/vector-store architecture for RAG products; own retrieval quality
- Set data quality, labeling, and validation standards for evaluation and fine-tuning datasets
- Build reusable schema-on-read patterns for semi-structured ERP exports (SAP, Oracle, NetSuite)

3. MLOps, Evaluation & Observability

- Build and own an evaluation/regression harness across all Lab products — accuracy, latency, cost-per-transaction, exception rate
- Govern prompt/model versions in production; roll back safely, catch provider-side regressions
- Instrument cost, latency,



and drift observability with alerting before issues reach pilots
- Define CI/CD for AI products — prompt testing, eval-gated deploys

4. Product Development & Prototyping

- Lead hands-on development of AI agent prototypes, Gate 0 to Gate 1
- Build and maintain agentic workflows on LLM APIs (Claude, OpenAI, open-source models)
- Integrate financial ERP systems (SAP, Oracle, NetSuite), reconciliation platforms, document processing APIs
- Ship production-quality prototype code, not proof-of-concept demos
- Instrument per-MVP evaluation metrics: accuracy, latency, cost-per-transaction, exception rate

5. Client Pilot Technical Support

- Deploy MVPs into client pilot environments — secure data handling, access management, audit logging
- Diagnose and resolve technical issues during active pilots with the Lab Head and delivery team
- Turn pilot findings into product backlog items
- Prepare technical due-diligence materials for client security/procurement reviews

6. People Leadership

- Hire and grow the tech team (AI/ML and full-stack) as the Lab scales, staying hands-on in code review and architecture decisions
- Own sprint cadence — planning, prioritization, demos — aligned to Gate milestones
- Run performance management and career development for direct reports, anchored to technical growth; flag skill gaps to the Lab Head
- Allocate work across AI/ML and full-stack to maximize throughput without compromising quality or security

7. Technology Governance & Risk

- Maintain a technology risk log: API dependencies, model performance, data handling, vendor concentration
- Own unit economics per product — API cost, infra cost, per-transaction margin
- Escalate client data usage, consent, or residency issues before go-live
- Ensure technology-partner arrangements are reflected in IP and vendor agreements

QUALIFICATIONS & EXPERIENCE

Essential

- 8–10 years of software engineering or ML engineering experience, including 4+ years in a technical leadership/team-lead capacity
- Production experience building and deploying LLM-powered applications or autonomous AI agents
- Robust Python; API integration patterns, async architecture, data pipeline design
- Built data pipelines for multi-modal/unstructured inputs — OCR, embeddings/vector stores for RAG, semi-structured extraction
- Integrated financial ERP APIs or document-processing pipelines (SAP, Oracle, NetSuite, or equivalent)
- Working knowledge of RAG architectures, prompt engineering, and LLM evaluation — has built evaluation/observability tooling, not just used it
- Demonstrable 0-to-1 build experience: concept to client-facing prototype
- Hired, managed, and developed a technical team spanning AI/ML and software engineering disciplines





Preferred

- Full-stack development experience (React/TypeScript or equivalent) sufficient to review and unblock front-end/dashboard work
- Experience in Finance & Accounting process automation (AP/AR, reconciliation, expense management, contract management)
- Exposure to agentic frameworks: LangChain, LangGraph, AutoGen, CrewAI, or similar
- Experience with data residency, SOC 2, DPDP 2023, or client data handling requirements in an enterprise SaaS or BPO context
- Prior work in an outsourcing, shared services, or BPO technology environment
- Experience with MLOps/LLMOps tooling (LangSmith, Weights & Biases, MLflow, or equivalent) for evaluation and observability at scale

Education

- Bachelor's or Master's degree in Computer Science, Engineering, or a quantitative field
- Relevant skilled certifications (AWS/GCP/Azure AI, Databricks, or equivalent) are advantageous but not required

COMPETENCY PROFILE

Builder Mentality

Comfortable with ambiguity; gets to a working artefact quickly and iterates. Produces code, not slides.

System Thinking

Designs for scale, security, cost, and observability from the start — not retrofitted after the demo.

Commercial Awareness

Understands unit economics. Tracks cost-per-transaction and ties architecture decisions to margin.

Client Orientation

Treats client data with production-grade rigor, including DPDP 2023 obligations. Zero tolerance for data-handling shortcuts.

Team Multiplier

Grows the team’s technical capability by staying hands-on: hires for real gaps, reviews code personally, unblocks rather than delegates. Not a people-manager-first profile.

Risk Clarity

Surfaces technical risks early and in plain language — does not manage upward by omission.

Data Engineering Rigor

Handles unstructured and multi-modal data (documents, images, semi-structured exports) with the same rigor as structured data. Data quality and lineage are first-order, not afterthoughts.

WORKING CONTEXT & STRUCTURE

The Engineering Manager, AI/ML Innovation & Product Engineering operates within a small team. The Lab is structured for speed and accountability — there are no large committees, no handoffs to a separate QA team, and no separation between 'build' and 'deploy.' The expectation is that this individual owns the technical outcome end-to-end.

Direct Reports

Lab tech team (AI/ML + full-stack engineers)

Reporting to

Innovation Lab Head (commercial lead)

Peer Relationships

BPS SMEs / Delivery Leads (domain input); US Hunter (client requirements translation)

Governance Interface

BPS CEO — escalation path for IP, data, and vendor risk issues

Technology Stack

Python, LLM APIs (Claude, OpenAI), agentic frameworks, financial ERP APIs, cloud infrastructure (AWS/GCP/Azure)

Key Dependencies

IP arrangement with technology partners must be resolved before development begins; client data consent framework must be in place before pilot deployment

Travel

Primarily India-based. Occasional travel to client sites or US for pilot support as required

📌 Artificial Intelligence Engineer (Pune)
🏢 Crescendo Global Leadership Hiring India
📍 Pune