Role | AI Solution Architect | Business AI Solutions & AI Platform Management |
Location | Chennai | Flexibility for business travel |
Experience | 12+ years in technology, including 8+ years in cloud-native AI/ML, GenAI or solution architecture |
What awaits you / Job Profile? | - Partner with business and product teams to frame business problems, define measurable outcomes and determine whether deterministic automation, predictive ML, GenAI, RAG or agentic AI is the appropriate solution.
- Design secure, scalable and cost-effective end-to-end AI solution architectures, covering experience, application, agent, model, data, integration, platform, security and operations layers.
- Own solution architecture from discovery and proof of value through production deployment, adoption and lifecycle management, including architecture decisions, non-functional requirements and technical risk.
- Build reference implementations and critical components; guide AI engineers and developers on Python, APIs, RAG, agent workflows, cloud-native patterns, evaluation and production-readiness practices.
- Design and manage reusable AI platform capabilities such as model gateways and catalogues, prompt and agent registries, vector services, orchestration, MCP/A2A connectivity, evaluation, guardrails, observability and chargeback/showback.
- Integrate AI solutions with enterprise systems, APIs, events, data platforms, identity services and knowledge sources while maintaining lineage, access control and auditability.
- Define LLMOps/MLOps and DevSecOps patterns for versioning, automated testing, deployment, monitoring, drift detection, rollback, incident response and continuous improvement.
- Embed Responsible AI, privacy, cybersecurity, intellectual-property protection, human oversight and regulatory controls into solution design and platform operations.
- Run architecture reviews and design gates; validate generated architecture and code, implementation alignment, resilience, performance, disaster recovery, security and cost efficiency.
- Define solution scorecards for task success, groundedness, safety, latency, reliability, adoption, cost per outcome and business value; use evidence to scale, redesign or stop solutions.
- Work with platform engineering, cloud, data, cybersecurity, enterprise architecture and product teams; communicate options, trade-offs and risks clearly to leadership and stakeholders.
|
What should you bring along? | - Proven delivery of enterprise AI solutions from business discovery and architecture through hands-on implementation, production operations and scaled adoption.
- Strong understanding of AI/ML, Generative AI and agentic systems, with practical judgment on model, data, integration, platform and build-versus-buy trade-offs.
- Deep cloud-native and distributed-systems foundation, including microservices, APIs, event-driven architecture, identity, resilience, observability and DevSecOps.
- Experience operating shared AI platforms for multiple products, including onboarding, tenancy, capacity, service levels, platform upgrades, cost governance and support processes.
- Ability to translate ambiguous business requirements into executable architecture, measurable acceptance criteria and incremental delivery plans.
- Strong communication, presentation, facilitation and negotiation skills, with the ability to influence product owners, engineers, architects and senior stakeholders.
- Bachelor’s or Master’s degree in Computer Science, AI, Data Science, Engineering or a related discipline, or equivalent professional experience.
|
Must-have technical skills | - Programming and application architecture: hands-on Python, REST/async APIs, microservices, event-driven patterns, structured outputs, automated testing and secure software engineering.
- GenAI solutions: model selection, prompt and context engineering, embeddings, vector and hybrid search, RAG, reranking, fine-tuning options, multimodal patterns and grounding/citation design.
- Agentic AI: planning and tool-use patterns, state and memory, orchestration, multi-agent design, human-in-the-loop controls, workflow recovery, sandboxing and least-privilege execution.
- Frameworks and protocols: practical experience with LangGraph/LangChain, Semantic Kernel, AutoGen or cloud-native agent services; MCP and A2A concepts, tool contracts and secure agent communication.
- AWS architecture: Bedrock and/or SageMaker plus core services such as Lambda, EKS/ECS, S3, RDS/Aurora, VPC, IAM, KMS, Secrets Manager, CloudWatch, API Gateway and event services; AWS Well-Architected principles.
- AI platform engineering: Docker, Kubernetes, Terraform/CloudFormation/CDK, CI/CD/CT, GitOps, model/prompt/agent registries, experiment tracking, artefact lineage and environment promotion.
- LLMOps/MLOps and observability: MLflow or equivalent, automated offline/online evaluation, tracing, token/latency/cost monitoring, data and model drift, canary or shadow release, rollback and incident management.
- AI evaluation and security: task-specific golden datasets, groundedness and hallucination checks, red-teaming, prompt-injection and data-exfiltration defenses, content safety and tool guardrails.
- Data and knowledge architecture: relational and NoSQL stores, vector databases, data lakes/lakehouses, Kafka or equivalent streaming, metadata, lineage, semantic models and knowledge graphs.
- Governance: Responsible AI, privacy by design, threat modelling and working familiarity with NIST AI RMF, ISO/IEC 42001, ISO 27001 and OWASP Top 10 for LLM Applications.
|
Good-to-have technical skills | - AWS Certified Solutions Architect – Professional and/or AWS Certified Machine Learning Engineer; equivalent Microsoft Azure or Google Cloud AI credentials.
- Experience with Amazon Bedrock Agents, Amazon Q, agent evaluation, model routing, vLLM/TGI, open-source models, GPU capacity planning and inference optimization.
- Experience building AI factories or internal developer platforms with reusable templates, model catalogues, automated policy checks and self-service deployment paths.
- Knowledge of advanced RAG, GraphRAG, knowledge graphs, ontologies, synthetic data, digital twins, edge AI or autonomous systems.
- Experience applying AI agents within the SDLC, including requirements, architecture, coding, testing, security remediation, release and operations.
- Automotive or another regulated-industry background with strong cybersecurity, privacy, traceability and audit expectations; FinOps and chaos-engineering experience.
|