AVP, GenAI Enterprise Architect at EXL

Job description

Design and deploy enterprise-grade AI solutions (LLMs, RAG, agents) by selecting appropriate models, building data pipelines, and integrating them with cloud platforms (AWS, Azure, GCP). Lead technical strategies across the standard 7-layer GenAI stack (from data ingestion to application interfaces), ensure scalability, manage AI security/hallucinations, and bridge business needs with engineering teams. • System Design & Architecture: Architect end-to-end Generative AI systems by operationalizing the 7-layer AI architecture (Data Sources, Preprocessing, Model Selection, Orchestration, Inference, Integration, and Application). • Model Selection & Tuning: Evaluate and select cutting-edge commercial (e.g., GPT-4) and open-source models, and fine-tune models for domain-specific use cases. • LLMOps, Observability & Pipelines: Establish LLMOps standards for model versioning and CI/CD. Implement foundational observability (OBS) layers using tools like Datadog, Splunk, or Prometheus to monitor system health, API latency, and basic application metrics. • Integration & Data Protection: Integrate AI solutions with existing APIs while enforcing core data protection measures, including Role-Based Access Control (RBAC), data encryption in transit, and basic PII (Personally Identifiable Information) masking to manage hallucinations and adversarial attacks. • Strategic Leadership: Collaborate with stakeholders to map business challenges to AI solutions and establish AI governance frameworks. • Client Consulting: Act as the primary onshore technical liaison, facilitating client workshops, requirements gathering, and translating business pain points into technical AI blueprints. • Consulting Skills: Exceptional client-facing communication skills; proven ability to present complex technical concepts to business stakeholders. • Technical Expertise: Deep knowledge of NLP, Python, deep learning frameworks (PyTorch/TensorFlow), and orchestration tools (LangChain, Autogen). • Cloud & Data Systems: Extensive hands-on experience with AI services on AWS, Azure, or GCP. Expertise in vector databases (e.g., Pinecone, Milvus) and embedding techniques. • Qualifications: Bachelor’s / Master’s in Computer Science, AI, Data Science, or related field; 8–15 years in software engineering, ML, or AI roles, with demonstrable onshore consulting experience. AVP, GenAI Enterprise Architect Experience Level: 8–15 Years | Location: Onshore Profile Focus: Hands-on AI Architecture, Pipeline Execution, & Client Delivery

Requirements

Must-haves

Model Selection & Tuning: Evaluate and select cutting-edge commercial (e.g., GPT-4) and open-source models, and fine-tune models for domain-specific use cases.
Strategic Leadership: Collaborate with stakeholders to map business challenges to AI solutions and establish AI governance frameworks.
Client Consulting: Act as the primary onshore technical liaison, facilitating client workshops, requirements gathering, and translating business pain points into technical AI blueprints.
Consulting Skills: Exceptional client-facing communication skills; proven ability to present complex technical concepts to business stakeholders.
Technical Expertise: Deep knowledge of NLP, Python, deep learning frameworks (PyTorch/TensorFlow), and orchestration tools (LangChain, Autogen).
Cloud & Data Systems: Extensive hands-on experience with AI services on AWS, Azure, or GCP. Expertise in vector databases (e.g., Pinecone, Milvus) and embedding techniques.
Qualifications: Bachelor’s / Master’s in Computer Science, AI, Data Science, or related field; 8–15 years in software engineering, ML, or AI roles, with demonstrable onshore consulting experience.
AVP, GenAI Enterprise Architect
Experience Level: 8–15 Years | Location: Onshore
Profile Focus: Hands-on AI Architecture, Pipeline Execution, & Client Delivery

Nice-to-haves

None flagged.

Application prep

Use the role’s language and application details while tailoring your materials.

Keywords to mirror

aiawsazuregcpci/cdapipythonml

Platform: oracle-cloud
Effort: medium