Ajith Reddy

Gen Ai Engineer @Leapgen Ai

Ashburn, VA, US

Led the development of Surround AI and Octopus, two GenAI platforms focused on enhancing RAG pipelines and domain-specific LLMs for enterprise use. Collaborated cross-functionally with product, engineering, and compliance teams to ensure model outputs met both technical standards and regulatory compliance- Built and deployed conversational AI solutions using Azure Bot Framework, Cognitive Services, and AI Studio. Fine-tuned GPT models (3.5 to 4) on Azure Databricks, maintaining version-controlled experiment logs and configurations aligned with Good Documentation Practices (GDP) and audit readiness- Created multimodal ETL pipelines using Azure Data Factory and Databricks to process unstructured data (PDFs, text, images), improving dataset quality by 35% and enrich ingestion workflows with Bing Search API, Logic Apps, and Power Automate, enabling structured indexing and reproducible results- Designed RAG systems with LangChain, LlamaIndex, and Azure Cognitive Search, improving contextual LLM performance by 40% while maintaining traceability through documented chaining logic and data lineage- Integrated content moderation filters via Azure OpenAI, reducing unsafe outputs by 60% with severity-level classifiers. Supported governance through moderation workflows and configurable filter logic documentation- Deployed services in Docker on AKS, boosting system reliability by 45% while implementing CI/CD pipelines using Azure DevOps and GitHub Actions for automated, traceable deployments- Continuously monitored system and model performance with Azure Monitor, Application Insights, and MLflow, using logs and feedback loops to guide optimization and compliance review- Analyzed user interactions using Azure Synapse Analytics to derive behavioral patterns and feed iterative updates by maintaining clear documentation of findings and alignment actions