Solution Architect - Skills & Reusability

Overview

Lead the technical design authority for enterprise AI and agentic solutions, setting reference architectures and making design decisions while partnering with product and engineering teams that build and operate solutions.

What you'll do

• Define and govern reference architectures and engineering patterns for AI, generative AI and agentic solutions.
• Lead technical design reviews and provide clear design pass, hold or remediation decisions.
• Design enterprise patterns for agent orchestration, retrieval and grounding, APIs, tools, skills and enterprise-system integration.
• Assess AI models for task suitability, quality, latency, security, economics and operational fit, and establish benchmarking approaches across model and workload classes.
• Drive token optimisation and cost-per-outcome engineering, including use of smaller, lower-cost models and routing strategies.
• Define evaluation, observability, LLMOps and model-drift patterns, including criteria to hold, roll back or promote solutions.
• Determine when model tuning is justified versus prompting, retrieval, tool and skill design or model selection.
• Build proofs of value for emerging technologies and convert successful findings into reusable enterprise patterns.
• Identify reuse opportunities before new engineering is commissioned.
• Work across platform boundaries, including cloud AI platforms, agent frameworks and deterministic automation technologies.
• Partner with Enterprise Architecture, security, privacy and engineering teams to ensure designs meet enterprise guardrails.
• Help mature solutions along the autonomy ladder, introducing human-in-the-loop, self-monitoring and automated remediation patterns.

What you'll need

• Strong enterprise or solution architecture experience, with hands-on depth in generative and agentic AI architecture.
• Proven experience designing production-grade agentic systems, including orchestration, tool and skill invocation, multi-agent patterns and human-in-the-loop controls.
• Strong grounding in large language models and model architecture, including model selection, benchmarking, context and token management and cost optimisation.
• Deep understanding of retrieval-augmented generation, enterprise search, embeddings, vector retrieval, grounding and knowledge integration.
• Strong knowledge of agent integration patterns, including APIs, event-driven integration and Model Context Protocol (MCP) style tool integration.
• Experience designing evaluation frameworks covering quality, groundedness, task completion, latency, safety and cost.
• Experience with LLMOps or generative AI operations, observability, model drift and production monitoring.
• Strong cloud architecture experience with at least one major hyperscaler, such as Microsoft Azure or Google Cloud Platform.
• Ability to evaluate native platform capability versus custom pro-code engineering based on business value and architectural fit.
• Strong security and Responsible AI architecture understanding, including identity, access control, data boundaries, auditability and guardrails.
• Ability to produce clear technical design documents, reference architectures, architecture decision records and engineering standards.
• Hands-on experience with enterprise AI ecosystems such as Microsoft Azure AI, Azure OpenAI and Microsoft Foundry; Google Cloud Vertex AI and Gemini; Anthropic Claude; and Microsoft Copilot Studio and adjacent enterprise agent platforms.
• Experience with agent orchestration frameworks such as Agent Development Kit (ADK), Semantic Kernel or equivalent.
• Strong stakeholder management, with the ability to challenge engineering teams and platform vendors constructively.

Nice to have

• Experience with robotic process automation (RPA) and broader hyperautomation architecture.
• Experience with prompt and context engineering.
• Knowledge of AI financial operations (FinOps) and model economics.
• Experience with automated testing for AI systems and continuous integration and continuous delivery patterns for prompts, agents and evaluations.
• Experience designing self-healing or NoOps-oriented automation patterns.
• Exposure to agent and skill catalogues, reusable components and platform engineering.
• Ability to translate emerging technology into practical enterprise adoption decisions: adopt, wait or stop.
• Experience in fast-moving consumer goods, retail and eCommerce, or global manufacturing organisations, with an understanding of complex global operations, high transaction volumes and supply chain processes.

Details

• Location: Bangalore Centre.