Keerthi Chillakuru

Constructed multi-agent orchestration system utilizing AI to enhance SOP querying and workflow insights.

• Created dedicated agents for efficient document retrieval, process analysis, compliance, and escalation.
• Deployed Supervisor Agent to manage workflows by routing queries and orchestrating agent functions.
• Implemented advanced extraction techniques with pattern matching and JSON parsing for SOPs.
• Leveraged Neo4j for representing SOP components and their interrelations to optimize query performance.
• Established robust state management features, including real-time tracking and error handling.
• Designed architecture for flexibility and scalability, accommodating future integrations of agents and tools.
• Facilitated automated analysis of SOPs, improving compliance validation and reducing manual workload.

• Developed an innovative solution aimed at optimizing supply chain networks by employing Generative AI technologies.
• Designed an Agent-Based Generative AI Framework: Created a dynamic AI agent that utilizes generative modeling to simulate and optimize various supply chain scenarios. This approach helps in predicting potential bottlenecks and efficiently managing resources.
• Constructing a Comprehensive Astra Database: Developed a detailed database to represent the intricate relationships and dependencies within the supply chain network. This database serves as the backbone for all AI-driven analyses and optimization tasks.
• Utilizing Lang Chain Agent and Tools: Integrated the Lang Chain agent to iteratively frame and refine queries for a large language model. This process is based on initial reasoning and dynamically generated responses, ensuring the derivation of practical and effective solutions.
• Accessibility for Non-Technical Users: Designed the interface and interaction model to be user-friendly, enabling non-technical personnel to inquire and receive optimizations through simple natural language inputs.