Azure & Claude AI Deployment Engineer – Python, Data Pipelines & Custom Plugins (Gurgaon / Remote)

# Full-Stack Contractor Specs: **Purpose**: Engagement spec for a full-stack contractor/contracting firm to support build of a multi-tenant enterprise SaaS platform. General architectural level -- no proprietary business logic, vendor/product names, industry vertical, or client details disclosed. --- ## System Archetype A multi-tenant, AI-agent-orchestrated SaaS platform where external enterprise clients log in and work through a structured, multi-step guided workflow. The platform ingests client-provided data alongside a separately maintained reference dataset, uses an AI orchestration layer to coordinate data retrieval and computation, and generates business documents (presentation and word-processing formats) as output. **Scale**: ~100 tenant organizations, ~50 users each (~5,000 total end users). Tenant data must be strictly isolated at the application layer. --- ## Architecture Layers & Required Skills | Layer | What it involves | Skills needed | |---|---|---| | AI agent orchestration | Backend service coordinating multi-step reasoning and tool use; manages guided-workflow session state | Backend development (Python or similar), LLM API / agentic tool-calling experience, API design, session-state management | | Data connectors | Structured data access exposed to the orchestration layer; tenant-scoped and shared reference-data access, both audit-logged | API/backend development, row-level security or scoped-token isolation patterns, audit logging | | Data transformation | Ingests a raw external feed and applies a proprietary methodology (logic withheld, provided separately under NDA); contractor builds pipeline/versioning infrastructure | ETL/data pipeline development, data versioning, secure storage architecture | | Human-in-the-loop mapping | User uploads a file; an automated process proposes a field mapping; user reviews/edits via confidence-scored, editable interface | Interactive frontend components (editable tables), integration with automated structured output, review-workflow UX | | Document generation | Programmatic generation of formatted business documents from computed results, callable from the orchestration layer | Programmatic document generation (standard office-format libraries), template-based rendering | | Client-facing UI | Guided, multi-step wizard interface on a governance-focused application platform (SSO/RBAC/audit logging built in); contractor implements screens and API integration | Experience with low-code/application platforms, REST API integration, SSO/RBAC familiarity | | Hosting & infrastructure | Backend hosted on a major public cloud, via managed or self-hosted compute; standard enterprise identity integration | Public-cloud infrastructure (compute, networking, identity), infrastructure-as-code | | Data governance / gateway | Centralized routing, auth, and access-policy enforcement across data connectors as they scale | API gateway configuration, access-policy/identity governance, centralized audit log aggregation | --- ## Deliverables Expected From Contractor(s) 1. Working code for the assigned layer(s), with documentation 2. Interface/API documentation for integration with other layers/contractors 3. Test coverage demonstrating tenant-isolation or mapping-accuracy correctness, as applicable 4. Deployment scripts/infrastructure-as-code (for hosting/infra work) 5. Knowledge transfer session(s) before contract close