Principal Architect - AI

Educational Qualification

• B.Tech / M.Tech / MS / Ph.D. in Computer Science, Information Technology, Artificial
• Intelligence, or a related technical field.
• Certifications preferred: AWS Solutions Architect Professional, Azure Solutions
• Architect Expert, or GCP Professional Cloud Architect.
• Research papers, case studies, or significant open-source contributions (are preferred)

Experience

• Minimum 12+ years of total professional experience.
• At least 5 years in AI/ML system design
• Proven experience designing and deploying multi-component AI solutions for government or enterprise environments.

Key Responsibilities

• Design and implement end-to-end solution architectures embedding AI/ML capabilities into NeGD and allied government applications.
• Define integration standards, APIs, and interoperability frameworks between AI systems and existing Digital India platforms.
• Oversee data flow architecture, security, privacy, and Responsible AI implementation across solutions.
• Architect and oversee integration of AI applications with NeGD's current enterprise applications and digital platforms.
• Evaluate and recommend cloud-native and on-premises deployment models in alignment with MeitY guidelines.
• Provide architectural leadership, ensuring scalability, maintainability, and compliance with MeitY and NIC technical standards.
• Collaborate with Data Science, MLOps, and Engineering teams for unified architecture governance.
• Conduct periodic architectural reviews and provide technical mentoring to development teams.

Technical Competencies:

• Architecture: Microservices and serverless patterns, event-driven and domain-driven design, high-level data & model lifecycle architecture
• Cloud: AWS (Bedrock, SageMaker, Lambda, API Gateway, S3), Azure (OpenAI Service, ML Studio, AKS), GCP (Vertex AI, Big Query, Cloud Storage), multi-cloud integration & cost optimization
• AI Frameworks: TensorFlow, PyTorch, Hugging Face Transformers, Lang Chain (LLM orchestration), prompt-engineering and responsible AI practices
• Data & Storage: SQL (PostgreSQL, MySQL), NoSQL (MongoDB, DynamoDB), vector databases (Pinecone, Weaviate) for RAG & semantic search
• APIs & Integration: REST, GraphQL, WebSocket for real-time chat/voice
• Conversational AI & Voice: Speech-to-Text and Text-to-Speech APIs (AWS Transcribe/Polly, Azure Speech), RAG/chatbot pipeline design with vector stores & LLMs
• Security & Governance: IAM and RBAC across clouds, encryption in transit & at rest, data privacy & AI governance (GDPR, SOC2)