SHANTHA LAXMI KUMAR

Generative AI Engineer/ Team Lead

Accenture

01.2021 - Current

Designed and deployed scalable, production-ready Gen AI solutions on Azure Cloud, including an LLM-powered system for client visit agendas that integrated seamlessly with existing business systems via serverless Azure Functions.
Managed the deployment and performance optimization of Gen AI APIs, implementing iterative prompting and validation strategies to ensure model reliability and consistent, high-quality output for business stakeholders.
Implemented and managed model monitoring and safety systems for Gen AI applications using Azure Content Safety and open-source frameworks (e.g., LLMGuard). This work addressed potential issues like performance degradation and ensured responsible AI deployment.
Engineered and productionized a text-to-SQL agentic AI chatbot, focusing on query optimization and infrastructure efficiency. Implemented RBAC to ensure secure, reliable data access within the ML system.
Led a team to a final-round victory in the Google Agentic AI Hackathon, prototyping a multi-agent solution on the Google Cloud Platform (GCP) using Vertex AI Search and Conversation (RAG), Gemini, and Cloud Run, demonstrating hands-on experience with modern GCP ML services.
Developed automated LLM evaluation frameworks (e.g., DeepEval) to systematically assess model performance and quality, creating a CI/CD-integrated testing loop that reduced manual validation efforts.
Collaborated with business stakeholders and technical teams to conduct PoCs (DALL-E, Azure OpenAI), identify high-value use cases, and prepare technical documentation on feasibility and implementation.
Multiple prototypes for multi-agent systems for requirements to QA test case generation, Playwright script generation, and execution with report analysis.
Mentored a team of junior engineers on AI/ML concepts, best practices for building reliable AI systems, and automation strategies.

Similar Profiles