Aman Kumar

About Aman Kumar

AI Engineer | Full-Stack Developer | Production-Oriented Intelligent SystemsI am an AI Engineer focused on building practical, production-oriented systems—not just models, but complete applications that integrate intelligence into real products.My work centers on generative AI and intelligent systems, including RAG-based architectures, semantic search, embeddings, and model-driven workflows. I enjoy designing the entire lifecycle of an AI system—from data ingestion and retrieval to reasoning and response—while prioritizing scalability and reliability.Core Technical FocusOn the engineering side, I primarily work with Node.js and the MERN stack to build AI-enabled backends and user-facing applications. I believe the model is only one part of a larger, well-designed system, and I specialize in shipping full-stack products that bridge that gap.Active Areas of Development• Retrieval-Augmented Generation (RAG): Designing robust pipelines for context-aware AI.• Vector Search & Embeddings: Implementing efficient semantic retrieval systems.• AI Automation: Building intelligent workflows that reduce manual friction.• End-to-End Implementation: Taking AI features from a local notebook to a production environment.My PhilosophyI approach AI engineering with a systems mindset—thinking beyond APIs to how AI features behave under real usage, evolving data, and production constraints. My background in full-stack and backend systems allows me to design solutions that are practical, maintainable, and scalable.For me, AI engineering is about turning intelligence into usable software, learning at pace, and building systems that solve real-world problems.

Experience

1. Product and Growth Intern

   Elvan

   Apr 2026 — Present

   Built and shipped AI-powered growth systems end-to-end for an early-stage SaaS startup (Elvan — elvan.ai):• Developed n8n agentic workflows for Reddit, X, and LinkedIn automation running on production scheduled triggers• Built Apollo.io enrichment pipelines processing 500+ leads with deduplication and SmartLeads delivery• Designed LLM pipelines (Gemini / GPT-4o mini) for intent detection, classification, and cold email personalization• Automated full outbound sequence — data scraping → enrichment → personalization → SmartLeads delivery• Built a CX Community Signal Monitor scraping Reddit via public JSON API, enriching with GPT-4o mini, writing to Notion• Contributed to Product Hunt launch (Vercel Day, April 2026) — GA4 tracking, ranking analysis, and social promotionTools: n8n · SmartLeads · Apollo.io · Claude · Gemini · GPT-4o · Notion

Education

• PSIT Kanpur (Pranveer Singh Institute of Technology)

  Bachelor of Technology - BTech, Computer Science

  2023 — 2027