Malay Chand

•Designed and deployed AI-powered calling agents (inbound & outbound) using prompt engineering, aligning agent behavior with client-specific workflows. •Built LLM-based analytics pipelines to process call transcripts and generate actionable insights for performance optimization. •Automated post-call analytics and funnel reporting, updating structured datasets with key business metrics. •Implemented tool-calling for prompt and TTS preprocessing, improving speech naturalness and conversational flow.

Key Projects & Contributions RAG-based AI Assistant: Designed a Retrieval-Augmented Generation (RAG) pipeline using LlamaIndex and OpenAI, optimizing chunking and embedding strategies to improve retrieval accuracy and GenAI performance. Python Executable for Repo Cleanup: Developed a .exe script to sort XML files, remove duplicates, flag tag inconsistencies, and detect anomalies with line-level warnings—eliminating reliance on ADX open-source tools. Repo Validation with CI/CD: Built YAML-based CI/CD pipelines to validate and clean Salesforce repositories by removing unused classes, profiles, and permission sets, significantly reducing pipeline failures. Apex Test Coverage Integration: Automated test coverage tracking using Salesforce API, mapping classes to their test cases and generating JSON coverage reports to enforce PR validation within pipelines. Email-to-Case Automation: Implemented automated case handling with dynamic queue assignment and auto-population of account/contact fields, improving support workflow efficiency. Deployment & Release Management: Managed CI/CD setup and bug fixes across SIT, UAT, Stage, and Production using Azure DevOps, ensuring smooth and reliable Salesforce deployments.