Principal Architect — Cloud, Agentic AI & Data Engineering

Principal Architect — Cloud, Agentic AI & Data Engineering

Field

Detail

Job Title

Principal Architect — Cloud, Agentic AI & Data Engineering

Function / Practice

Digital, Data & AI — Technology & Solutions

Reports To

Head of Architecture / Practice Head — Digital & AI

Experience

15+ years in software, data and cloud engineering, including 5+ years in an architect capacity

Location

Hyderabad / Bengaluru / Pune (Hybrid), with travel to client sites as required

Employment Type

Full-time, Permanent

Grade / Band

Principal Architect

ROLE SUMMARY

Cyient is looking for a Principal Architect who can operate at the intersection of cloud platform engineering, agentic AI and modern data engineering — and who is equally comfortable in front of a customer as in a design review.

This is a hands-on, customer-facing architecture role. You will own end-to-end solution architecture for strategic engagements, shape winning proposals and RFP responses, build demos and proofs of concept alongside the delivery team, and set the technical direction that engineering teams execute against. You will also mentor architects and senior engineers, and partner with fellow architects across practices to drive reusable assets, accelerators and an active innovation agenda.

Success in this role looks like architecture decisions that survive production, customers who ask for you by name, proposals that convert, and a bench of engineers who are measurably stronger because you invested in them.

KEY RESPONSIBILITIES

Solution & Cloud Architecture

• Own end-to-end architecture for large, multi-workstream programs spanning cloud infrastructure, data platforms and AI services.
• Design cloud-native solutions on Microsoft Azure (or AWS ) covering compute, networking, identity, integration, observability, cost management and security.
• Define the non-functional architecture — scalability, resilience, latency, availability, disaster recovery and multi-tenancy — and hold designs accountable to it.
• Establish architecture governance reference architectures, design standards, review checkpoints, a technology radar and architecture decision records (ADRs).
• Drive cloud modernisation and migration strategy — landing zones, containerisation, microservices, event-driven patterns and infrastructure-as-code (Terraform / Bicep / ARM).

Agentic AI & Applied GenAI

• Architect agentic AI systems — multi-agent orchestration, tool and function calling, planning and reflection loops, memory, and human-in-the-loop control points.
• Select and integrate the AI stack LLM providers (Azure OpenAI, Anthropic Claude, open-weight models), orchestration frameworks (LangGraph, Semantic Kernel, AutoGen, MCP-based tooling) and vector / retrieval layers.
• Design production-grade RAG and knowledge-grounding pipelines — chunking strategy, embeddings, hybrid search, re-ranking, grounding and citation.
• Define AI evaluation and guardrails accuracy and hallucination benchmarks, red-teaming, prompt-injection defence, PII handling, cost and token governance, and model observability.
• Operationalise LLMOps — prompt versioning, CI/CD for AI assets, regression evaluations, drift monitoring and safe rollout patterns.
• Champion Responsible AI principles — transparency, traceability, bias assessment and auditability — aligned to client and regulatory expectations.

Data Engineering & Platform

• Architect modern data platforms — lakehouse and medallion architectures on Azure Databricks, Microsoft Fabric, Synapse, Snowflake or equivalent.
• Design batch and streaming pipelines using Spark, Kafka / Event Hubs, Azure Data Factory and dbt, with clear SLAs on freshness and data quality.
• Establish data governance cataloguing, lineage, master data management, quality frameworks and access control (Unity Catalog, Microsoft Purview).
• Model data for both analytics and AI consumption , ensuring the same governed assets serve BI, ML and agentic workloads.
• Optimise for cost and performance — partitioning, storage tiering, compute right-sizing and workload isolation.

Customer Engagement & Presales

• Act as the senior technical face to the customer — run discovery workshops, architecture reviews and executive briefings for both engineering and CxO audiences.
• **Translate business problems into solution opti…