Mihir Kapile

Full Stack AI Product Engineer with 2+ years of experience building and scaling production-grade AI systems that combine agentic workflows, LLM reasoning, and cloud-native architectures. Proven track record of owning end-to-end ML product development—from data ingestion and model orchestration to backend services, asynchronous execution, and deployment. Deep expertise in agentic AI architectures (LangGraph, AutoGen, CrewAI), RAG, and multi-step ML pipelines powering real-world SaaS platforms.

Strong full-stack background across Python, TypeScript, Java, and React, with hands-on experience deploying and operating ML systems on AWS using Docker and Kubernetes.

Full Stack ML Product Engineer at Ease Vertical AI (2025-11 – Present)

• Led end-to-end development of EaseGTM™, an AI-powered SaaS go-to-market automation platform, owning system design across ML orchestration, backend services, data pipelines, and cloud infrastructure.
• Architected dependency-aware agentic pipelines using AutoGen, enabling multi-stage GTM generation with downstream reasoning dynamically adapting to upstream outputs.
• Designed and implemented scalable LLM orchestration workflows with robust failure handling, retries, and state management, increasing the reliability of complex multi-step AI executions in production.
• Optimized asynchronous task execution by refactoring Celery-based orchestration, enabling concurrent multi-user workloads and improving platform throughput, latency, and horizontal scalability.
• Built data ingestion and preprocessing pipelines for structured and unstructured inputs (PDFs, Excel, proprietary documents), embedding and indexing data into Weaviate vector databases to power RAG
• Translated customer domain expertise and proprietary business data into execution-ready AI workflows, reducing reliance on manual spreadsheets and disconnected tools while accelerating GTM strategy creation.
• Improved platform performance and operational stability by identifying bottlenecks across LLM calls, async processing, and vector retrieval, driving measurable gains in execution speed and system reliability.
• Deployed and operated the ML platform on AWS (ECS, S3) using Dockerized services, ensuring secure, scalable, and production-grade infrastructure for customer-facing AI workflows.

AI Engineer at Smartshore Services LLC (2025-06 – 2025-10)

• Contributed to the development of an Agentic AI-powered Source-to-Contract (S2C) platform as part of a cross-functional team. Helped design and implement multi-agent workflows for vendor discovery, onboarding, and RFP automation using LangGraph and CrewAI
• Built Agentic AI workflows to automate vendor identification, RFP creation, distribution, and RFP response evaluation using LLM-integrated logic.
• Developed intelligent vendor ranking and recommendation logic, enabling automated vendor onboarding.

Software Engineer at IpserLab LLC (2023-06 – 2025-05)

• Engineered a full-stack platform (React, NodeJS, MongoDB) enabling universities and enterprises to commercialize research papers and patents, enhancing visibility and revenue opportunities.
• Implemented a Java-based web scraper to extract metadata from academic and corporate sites systematically, surfacing research trends and institutional requirements to support lead targeting.
• Developed a modular website builder using Spring Boot, ReactJS, and MySQL, empowering clients to configure and launch tailored institutional portals without technical expertise.
• Devised and integrated a customizable AI chatbot that guided users through portal creation and addressed research-specific queries, reducing onboarding time and support dependency.

MS in Computer Science – Northeastern University, Khoury College (2024-05)

Bachelor of Technology in Computer Science and Engineering – MIT ADT University (2022-07)