Subhajit Mandal

AI/ML Engineer

HP • Tiempo completo

Feb 2022 - Present • 4 yrs 3 mos

• Architected and deployed an end to end Voice AI Agent integrating Vapi with Dentally CRM using FastAPI, Webhooks, and MCP (Model Context Protocol), enabling fully automated 24/7 appointment booking with zero manual intervention. • Designed 8+ REST API endpoints enabling real-time backend workflows: appointment search, slot booking, patient data updates, and dentist schedule validation all triggered autonomously by the Vapi voice agent. • Implemented production-grade security with HMAC webhook verification, raw body logging, and asynchronous request processing, ensuring reliable and secure communication between voice AI and backend services. • Built a conversation to action mapping engine that translates natural language voice interactions into structured API calls, reducing booking errors and eliminating 100% of manual scheduling overhead. • Deployed containerized FastAPI application to AWS EC2 with Nginx reverse proxy + Uvicorn workers, HTTPS, static IP exposure, and automated environment handling achieving 99.9% uptime. • Designed agent based workflows integrating LLM reasoning with backend APIs to automate real world business operations across the entire patient lifecycle. • Integrated OpenAI Whisper for real time speech to text transcription in voice agent pipelines, enabling accurate natural language understanding across diverse accents and noisy environments. • Built a Custom Voice Agent using OpenAI Whisper TTS/STT and Anthropic Claude for intelligent, context-aware conversational interactions with low-latency response. • Leveraged OpenClaw to automate entire business processes including marketing, lead generation, content creation, and multi platform social posting (LinkedIn, Instagram), dramatically reducing manual effort across client campaigns. • Developed Content Generation & Content Posting AI Agents and Lead Generation AI Agents using OpenClaw's multi-model agentic workflow to run autonomous marketing and sales pipelines.

Software Engineer

Amazon • Tiempo completo

Aug 2020 - Feb 2022 • 1 yr 6 mos

• Performed data analysis on customer complaints and product performance data to identify root causes, trends, and recurring patterns, informing product quality improvements. • Delivered actionable insights to stakeholders through data driven reports, contributing to measurable reductions in complaint resolution time. • Do Full Stack Web Development