AI Solution Architect

Location: Pune, India

Job Requirements Please enter AI Solution Architect Agentic AI | Enterprise AI Architecture | Modern Data Platforms | Customer Solution Leadership Function AI Architecture & Intelligent Solutions Level Principal Architect / AI Solution Architect Experience 12-18 years overall, including 4+ years leading AI/ML or GenAI architecture Work model Client-facing / Hybrid, with on-ground engagement as required Role Purpose The AI Solution Architect will lead customer-facing AI programs from business problem definition through production adoption. The role will shape practical solutions across predictive AI, Generative AI, RAG, multimodal and agentic AI, while ensuring that every solution is grounded in trusted data, strong data-engineering practices and modern platforms such as Databricks and Snowflake. The architect will serve as solution lead, design authority and customer thought leader, balancing innovation, business value, security, responsible AI, scalability and cost. Key Roles & Responsibilities Business Problem Solving & Customer Advisory Lead discovery workshops with business and technology stakeholders to identify high-value problems, workflows and measurable outcomes. Determine whether the right solution is analytics, classical ML, GenAI, RAG, agentic AI or a combination, and explain trade-offs in business language. Convert selected use cases into solution blueprints, value hypotheses, roadmaps, estimates, risks and acceptance criteria. Communicate architecture, dependencies, costs and risks clearly to customer executives, product owners and engineering teams. AI & Agentic Solution Architecture Design secure, scalable architectures for ML, GenAI, RAG, copilots, multimodal and single/multi-agent solutions. Define agent roles, orchestration, context, memory, planning, tool use, permissions, autonomy boundaries, human approvals, fallback and audit trails. Design access-aware retrieval covering ingestion, parsing, chunking, metadata, embeddings, vector/hybrid search, reranking, citations and freshness. Integrate AI solutions with APIs, events, enterprise applications, knowledge systems and systems of record while preserving identity and authorization. Apply model and platform patterns using Azure AI Foundry/OpenAI, Databricks Mosaic AI, AWS Bedrock/SageMaker, Google Vertex AI or equivalent services. Data Architecture & Modern Data Stack Ensure AI solutions are grounded in governed, high-quality structured and unstructured data, not isolated proof-of-concepts. Understand and guide data ingestion, ETL/ELT, batch/streaming, CDC, medallion architecture, lakehouse/warehouse, dimensional modeling, data products and semantic layers. Provide architecture guidance across Databricks, Delta Lake, Unity Catalog, MLflow/Mosaic AI, Snowflake, Snowpark, Cortex AI, dbt and cloud-native data services. Define data quality, metadata, lineage, access control, privacy, retention, observability and cost-management requirements for analytics and AI. Partner with data architects and engineers to design reusable pipelines, feature/context services and certified data products for AI consumption. Responsible AI, Evaluation & Productionization Define responsible AI controls covering privacy, safety, fairness, transparency, intended use, human oversight and incident response. Establish evaluation for task quality, groundedness, retrieval relevance, safety, robustness, latency, cost and business outcomes. Define MLOps/LLMOps practices for experiments, model/prompt/version management, CI/CD, monitoring, rollback and retirement. Lead threat modeling and guardrails for prompt injection, data leakage, unsafe tool execution, hallucination and uncontrolled agent actions. Ensure prototypes are hardened for security, reliability, observability, scalability and enterprise support. Technology Strategy & Thought Leadership Track and assess market developments in agentic platforms, reasoning models, multimodal AI, context engineering, open integration standards, small language models and AI-native data platforms. Separate durable enterprise capabilities from short-lived trends through build-vs-buy, platform and model evaluations. Develop customer points of view, reference architectures, accelerators, playbooks and executive briefings. Provide design authority, mentor AI/data engineering teams, support proposals and guide delivery from proof of value to scaled adoption. Mandatory Skills & Experience 12-18 years of technology experience, including 4+ years leading AI/ML or GenAI architecture and enterprise delivery. Deep knowledge of ML, LLMs, GenAI, RAG, embeddings, vector databases, multimodal AI, agent orchestration and tool-using agents. Hands-on architecture experience with at least one major AI ecosystem: Azure AI Foundry/OpenAI, Databricks Mosaic AI, AWS Bedrock/SageMaker or Google Vertex AI. Strong data fluency covering data engineering, SQL/Python, ETL/ELT, medallion/lakehouse architecture, dimensional modeling, semantic layers, governance and data quality. Working architecture knowledge of Databricks and Snowflake, including how their data, governance, analytics and AI capabilities fit enterprise solutions. Experience with APIs, microservices, event-driven integration, cloud architecture, identity, network security and enterprise application integration. Practical experience with AI evaluation, MLOps/LLMOps, observability, responsible AI, security controls and production release criteria. Proven ability to solve ambiguous business problems, lead customer workshops, influence senior stakeholders and communicate complex trade-offs clearly. Experience guiding cross-functional teams of AI engineers, data scientists, data engineers, software engineers, security specialists and product leads. Preferred Skills & Certifications Experience with Semantic Kernel, LangGraph/LangChain, LlamaIndex, AutoGen or comparable agent and orchestration frameworks. Experience with Databricks Vector Search, Azure AI Search, Snowflake Cortex Search, Pinecone, Weaviate, pgvector or similar retrieval technologies. Knowledge of dbt, Kafka, Airflow, Kubernetes, containers, GPU infrastructure, model serving and open-source model deployment. Understanding of Model Context Protocol (MCP), agent-to-agent interoperability, context engineering and deterministic workflow controls. Relevant Azure, Databricks, Snowflake, AWS or Google Cloud AI/data architecture certification. Consulting, pre-sales, RFP response, estimation, regulated-industry or practice-building experience. Expected Outcomes Customer-approved AI solution architecture, roadmap and architecture decisions tied to measurable business outcomes. Production-ready RAG/agentic, data, integration, security, evaluation and operating designs. Reusable reference architectures, guardrails, evaluation assets and delivery patterns that accelerate future AI programs. Clear customer guidance on technology choices, emerging trends, adoption risks and the path from pilots to governed scale. Equal Opportunity Selection decisions will be based on role-relevant capability, experience and demonstrated behaviors. Reasonable accommodations will be provided where required. Job descrption here....