Directly focused on building and architecting agentic LLM systems and enterprise AI agents (Vertex AI, agent frameworks).
About the Role
Lead the design and delivery of enterprise-grade agentic AI solutions, defining architectures, evaluation, and production requirements for autonomous AI agents using LLMs and cloud-native platforms. Collaborate with stakeholders to translate business processes into measurable AI-driven outcomes and guide engineering teams through prototyping to production.
Job Description
Role
Agentic AI Architect responsible for designing, architecting, and guiding implementation of agentic AI systems and intelligent agents for enterprise use. The role focuses on translating business processes into measurable outcomes, defining agent boundaries, safety controls, evaluation strategies, and production requirements.
Key Responsibilities
- Analyze current business processes, users, systems, handoffs, exceptions and failure points to determine where agentic AI is appropriate versus conventional solutions.
- Define target processes showing agent participation, deterministic controls, and human review/approval points.
- Convert business goals into measurable KPIs (baseline, targets, measurement methods, reporting cadence, owners).
- Design end-to-end architectures: models, RAG, knowledge sources, retrieval, agent orchestration, tools, APIs, state, memory, channels, identity and permissions, and human escalation.
- Define autonomy boundaries (read-only, confirmation-required, never delegated) and safety controls (privacy, tenant separation, audit evidence, prompt-injection resistance, least privilege).
- Specify evaluation strategy: representative scenarios, test datasets, acceptance thresholds, safety and regression gates, tool-action checks.
- Define production requirements: response time, availability, throughput, cost, logging, tracing, monitoring, incident handling, rollback, and update/change procedures.
- Select models, retrieval approaches and frameworks based on quality, security, latency, cost and portability.
- Build or guide technical prototypes to validate high-risk assumptions and review engineering artifacts (code, prompts, tests, data flows).
- Produce architecture diagrams, decision records, risk assessments, cost models, roadmaps and operating-model documentation.
- Provide technical direction to AI, QA, data, application, security and operations teams and track KPI outcomes post-deployment.
Requirements
- 10–15 years of experience in software, data, cloud or solution architecture with substantial hands-on production experience in AI/LLM systems.
- Deep knowledge of LLMs, prompting/structured generation, embeddings, vector/hybrid search, reranking, RAG, tool integration, agent orchestration, state and memory.
- Demonstrable expertise in at least one production AI stack (examples: Vertex AI/Gemini with ADK, LangChain/LangGraph, Semantic Kernel, LlamaIndex or equivalent).
- Strong ability to review Python code, APIs, data pipelines, retrieval logic, agent workflows, automated tests and deployment designs.
- Experience integrating with enterprise applications, APIs, data platforms, identity systems and approval workflows.
- Expertise in AI evaluation (answer quality, source support, tool-call correctness), safety, latency and cost considerations.
- Experience designing cloud-native, containerized or serverless solutions with CI/CD, secrets management, IAM, observability and rollback controls.
- Knowledge of responsible AI, privacy, secure AI architecture, audit and controls for action-capable systems.
- Experience in regulated environments with strict audit, privacy, residency and approval requirements is preferred.
- Relevant cloud, ML, generative AI or architecture certifications are desirable.
- Current and target process views documenting human, system and agent responsibilities.
- AI suitability assessments and capability statements (what the solution will/may/will not do).
- KPI definitions with baseline, target, measurement source and accountable owner.
- End-to-end architecture and data-flow diagrams with trust and permission boundaries.
- Evaluation, security, observability, cost and production-support strategies and phased delivery roadmap.
- Architecture decisions, identified risks, assumptions and dependencies.
- Competitive salary and benefits package.
- Quarterly growth opportunities and company-sponsored higher education and certifications.
- Employee engagement initiatives (project parties), flexible work hours, Long Service awards.
- Insurance: group term life, personal accident, Mediclaim hospitalization (covers self, spouse, two children, and parents).
- Support for hybrid work, flexible hours, accessibility-friendly offices with ergonomic setups and assistive technologies.
Location & Work Model
- Location: Hyderabad & Bengaluru (India). Role is full-time and supports hybrid work and flexible hours.
Vertex AI Gemini ADK Vertex AI Search LangChain LangGraph Semantic Kernel LlamaIndex Python BigQuery Cloud Storage Cloud Run GKE Pub/Sub Cloud Logging Cloud Monitoring Cloud Trace GCP RAG Embeddings Vector search Hybrid search Retrieval approaches Knowledge graphs Document intelligence Multimodal systems MCP CI/CD IAM
Skills
Solution Architecture System Design Agent Orchestration AI Evaluation Prompting / Prompt Engineering Cloud Architecture Security & Privacy Data Architecture Business Process Analysis Stakeholder Communication Workshop Facilitation Leadership Technical Writing Risk Assessment CI/CD and Release Management Observability and Monitoring Project Delivery QA/Test Strategy
Experience Level
Senior
Employment Type
Full Time, Permanent
- Competitive salary and benefits package
- Quarterly growth opportunities
- Company-sponsored higher education and certifications
- Project parties / employee engagement initiatives
- Flexible work hours
- Mediclaim hospitalization coverage (self, spouse, two children, parents)
- Hybrid work support
- Accessibility-friendly office with ergonomic setups and assistive technologies