About the groupThe Procurement Technology & AI Products group builds the procurement analytics and AI capabilities deployed across the firm’s operations, sourcing and transformation engagements. Its portfolio is used by consulting teams in live client work and is increasingly licensed to clients directly.

The portfolio today comprises seven products operating on a single governed data platform: a governed client spend cube and its onboarding pipeline; an AI analyst that converts a raw spend file into governed diagnostics and a narrated fact base; a peer-benchmarking agent over an anonymised consortium dataset, available in a web console and inside Microsoft Teams; commodity-strategy and should-cost engines; an evidence-based operating-model assessment with a frozen, auditable scoring engine; and a real-time voice interviewer that scores responses as the conversation unfolds. The platform is built on Azure and Azure Databricks, with large-language-model capabilities from OpenAI, Azure OpenAI and Anthropic in production.

Role summaryThe Senior AI Engineer is accountable for the production integrity of the product portfolio and the shared data platform on which it runs. The role combines hands-on engineering with technical leadership: defining the architecture, standards and operating practices that allow a compact team to run client-facing systems reliably, securely and at predictable cost, and leading the platform through its next phase of consolidation, hardening and scale.

This is a senior individual-contributor position with significant influence over engineering direction. The successful candidate will have owned production systems end to end and will bring that discipline to a portfolio that has grown quickly and now requires a common operating model.

Key accountabilitiesProduction operations and reliability· Define and maintain service-level objectives, health monitoring, alerting and incident-management practice across all products; lead incident response and post-incident review.

· Establish release management, environment promotion and rollback procedures so that every change reaches production through the pipeline, with auditable provenance.

· Own capacity, performance and cost management for application services, data warehouses and model consumption.

Platform architecture and data foundation· Steward the shared Azure Databricks workspace as a governed, multi-tenant foundation: catalog and schema design, Unity Catalog governance, warehouse and job strategy, secret scopes and cost controls.

· Own the contract layer through which products bind to the spend cube, and the onboarding path that takes a new client from raw files to a governed, queryable dataset.

· Lead the completion of the Snowflake-to-Databricks migration, including parity validation against legacy production,



which remains read-only throughout.

· Rationalise application state and integration patterns across products into a small number of sanctioned, documented approaches.

Security, identity and governance· Implement a single identity and authorisation model across the portfolio (Microsoft Entra ID), consistent API gateway policy (Azure API Management) and a uniform secrets posture (Azure Key Vault; no credentials in code or configuration).

· Complete the hardening programme: upload scanning and attestation, gateway rate and quota policies, network isolation, dependency and image hygiene, audit trails for regulated outputs.

· Ensure client data boundaries are enforced by design, including the principle that language models operate only over governed aggregates and never over raw client records.

Applied AI engineering· Set the engineering standard for LLM-backed features in production: evaluation datasets, regression testing, guardrails, tracing, provider failover and cost and latency budgets.

· Operationalise the designed LLM gateway (token accounting, caching, policy) and bring model usage under central observability.

· Partner with data scientists to move prototypes to production-grade services with clear ownership and runbooks.

Engineering practice and leadership· Define and enforce standards for code review, testing, infrastructure as code, documentation and on-call readiness; raise the baseline across the team through review and mentoring.

· Provide technical assessment of product requests from partners and clients, distinguishing clearly between capabilities that are in production, designed, or demonstrable.

· Represent engineering in architecture, security and vendor discussions within the firm.

Measures of success· Within 90 days: a production operations baseline in place — service health, job success, warehouse and model cost, and alerting — with documented ownership for each product.

· Within six months: all scheduled workloads deployed from configuration; the hardening backlog closed; application state migrated to governed, backed-up stores with tested restore; regression coverage for the spend-cube ELT on every merge.

· Within twelve months: the migration complete with parity evidence; a single identity, gateway, secrets and deployment model across the portfolio; a documented onboarding path that takes a new client to a governed dataset in days rather than weeks.





Required qualifications and experience· A minimum of eight years’ professional software engineering experience, including demonstrable accountability for production systems: service ownership, incident management, capacity and change management.

· Expert-level Python; professional proficiency in TypeScript and Node.js; the ability to review and contribute across both.

· Substantial experience delivering on Microsoft Azure: App Service or container platforms, storage services, Key Vault, Entra ID application registrations and token flows, API Management, virtual networking, and infrastructure as code (Bicep or Terraform).

· Production experience with Databricks or an equivalent lakehouse platform: Delta Lake, Unity Catalog or comparable governance, SQL warehouses, job orchestration and cost management; advanced SQL.

· Design and ownership of CI/CD pipelines, including environment strategy, secrets management and rollback.

· Applied security competence: OAuth 2.0 and OpenID Connect, least-privilege design, short-lived credentials, input validation, dependency management and audit logging.

· Delivery of at least one large-language-model capability relied upon by end users in production, including its evaluation approach, cost and latency management and failure handling.

· Data engineering judgement: medallion architectures, idempotent ELT, schema contracts and parity testing.

· Transparent, structured written and verbal communication with technical and non-technical stakeholders.

Preferred qualifications· Experience leading a warehouse-to-lakehouse migration (Snowflake to Databricks or comparable).

· Microsoft Bot Framework, Teams applications, Adaptive Cards and Microsoft Graph.

· Real-time audio streaming (WebRTC, WebSocket) and speech-to-speech model integration.

· Document AI: OCR, layout analysis and structured extraction from PDFs and presentations.

· Front-end proficiency in Next.js or Vue and an appreciation for information-dense analytical interfaces.

· Cloud and model FinOps.

· Domain knowledge in procurement or supply chain: spend analytics, category management, should-cost modelling, benchmarking, sourcing events.

· Experience in professional services or another environment in which delivery is paced by client commitments.

Technical environment· Languages and frameworks: Python (FastAPI, Flask), TypeScript/Node.js (Express, Next.js, Vue 3).

· Platform: Azure App Service, Azure Container Registry, Azure Functions, Azure Storage (Blob, Table, Queue), PostgreSQL, Redis, Key Vault, API Management, Entra ID, Bot Service.

- · Data: Azure Databricks (Unity Catalog, Delta Lake, serverless SQL warehouses, Jobs, Asset Bundles), Snowflake (legacy, read-only), MySQL control plane.

📌 Senior AI Engineer (Gurugram)
🏢 Kearney
📍 Gurugram