Engineering Lead (AI & Automation Products)

Lead engineering for AI and automation products under the Merkle brand, guiding a team of junior and mid-level developers based in India. The role is listed in Bengaluru and India - Maharashtra- Mumbai - Thane. It is full-time and permanent; no working arrangement or experience minimum is stated.

Technical architecture and delivery:

• Own architecture decisions across the portfolio, including database schemas, API contracts, integrations and workflow orchestration. Choose between direct API integration and Model Context Protocol (MCP) based on latency, control, security, auditability and reuse. Account for scope, timelines and quality across workstreams in partnership with the Director responsible for product requirements and prioritization.
• Design the PostgreSQL schema for pipeline data, access control, cost attribution and audits. Own the Claude/API integration layer, including system prompts, testing, ingestion of XLSX, DOCX, PPTX and PDF files, structured-output parsing, schema validation and fallback handling.
• Design multi-step AI-assisted workflows with background jobs, queues, retries, idempotency, human review, run state and replayable audit histories. Build DSP integrations such as DV360 and TTD, including OAuth and phased read/write access. Own credential management, multi-tenant permissions, client-scoped visibility and usage reporting by client and run.

Team leadership and implementation:

• Manage hiring input, onboarding, performance and growth planning. Conduct code reviews, pairing and mentoring; establish standards for testing, validation, AI evaluation, observability and code quality. Lead sprint-level technical planning, unblock the team and escalate cross-team blockers.
• Build Python APIs using FastAPI or similar tools and React/Tailwind front ends when needed. Use Claude Code daily and establish shared team conventions for its use. Instrument prompt and model traces, output validation, latency, costs and failures. Maintain AI evaluation harnesses with golden test sets, versioning, regression checks, failure taxonomies and release gates for higher-risk automations. Contribute reusable components and shared platform services.