Architect and govern enterprise-grade GenAI solutionsincluding RAG pipelines, intelligent agents, evaluations, and fine-tuning frameworks. This role ensures high-quality technical delivery, alignment with business needs, and the establishment of architecture standards for the AI Division.
- Accountability & Responsibilities
- Design end-to-end GenAI architectures (RAG, agents, LLM serving, orchestration).
- Evaluate use cases and translate them into scalable architecture blueprints.
- Define data, model, and integration architecture for GenAI workloads.
- Ensure performance, reliability, and governance of all AI solutions.
- Set architecture principles, coding standards, and MLOps patterns.
- Conduct design reviews, risk assessments, and quality checks on all AI deliveries.
- Oversee vendor/partner solutions to ensure compliance with internal standards.
- Provide technical direction to delivery teams (data engineers, ML engineers, developers).
- Own the technical success of GenAI projects and PoCs.
- Manage architecture documentation, diagrams, and decision records (ADRs).
- Align with business, product, and IT security teams to ensure business-ready solutions.
- Communicate risks, trade-offs, and solution options clearly to non-technical leaders.
- Evaluate emerging LLMs, tooling, and architectures.
- Lead experimentation, benchmarks, and performance evaluations.
- Drive continuous improvement of the AI platform and reusable GenAI components
Requirements
1 Required Experience
- 36 years in AI/ML engineering, software architecture, or distributed systems.
- Proven leadership of complex AI/GenAI solution deliveries.
- Experience in enterprise or regulated environments.
2 Technical Skills
- RAG Design & Evaluation
- LangChain / LlamaIndex
- Vector Databases: Qdrant, Milvus, FAISS
- LLM Serving: vLLM, Triton, TorchServe
- Guardrails: evaluation, safety, observability
- Event-Driven & Microservices Architecture
- Python, Node.js, FastAPI
- Containerization: Docker, Kubernetes
- MLOps & Experiment Tracking: MLflow, Weights & Biases
3 Tools & Platforms
- Python + Poetry
- FastAPI or Node.js
- Docker, Kubernetes
- MLflow, Weights & Biases
- GitHub Actions / CI/CD
- Cloud AI platforms (Azure OpenAI, AWS Bedrock, GCP Vertex AI)
4 Certifications
- NVIDIA NCA / ACE (Preferred)
- Cloud AI Architect Azure / AWS / GCP (Plus)
