Hi, my name is
Ankit Kumar.
I build high-performance web architecture.
Senior Software Engineer 2.2 at Bluecore with 7+ years of experience. I specialize in building scalable frontend systems, JavaScript engines, and AI-powered product features for Fortune 500 marketing brands.
About Me
I'm a Senior Software Engineer 2.2 at Bluecore, where I build and maintain enterprise-grade marketing platforms for Fortune 500 brands. Over the past 7 years, I've worked across the full stack — from architecting Vanilla JS engines for Tier-1 retailers like Nike and Reebok to building AI-powered tools that convert natural language into runnable audience segments.
My expertise lies in JavaScript internals, React, TypeScript, and Next.js, with deep experience in building high-performance frontend systems. I'm passionate about bridging the gap between complex engineering and business goals, leveraging Python and Agentic AI to drive intelligent automation and product growth.
Experience
Senior Software Engineer 2.2 · Bluecore
Leveraging AI (Claude Code) to automate data extraction across 450+ repositories for Python 2 to 3 migration. Architecting new user-facing features and optimizing legacy modules for improved performance and engagement.
Senior Software Engineer 2.1 · Bluecore
Enhanced the Audience Builder with sophisticated marketing journey tools. Engineered an AI Agent to convert natural language into runnable audience segments. Built a self-improving RAG database and an agent engine service with auth.
Software Engineer 1.2 · Bluecore
Scaled a Vanilla JS engine supporting complex campaign types including dynamic product recommendations and social proofing. Replaced jQuery with a custom library, achieving 5x bundle size reduction (30KB to 6KB).
Software Engineer 1.1 · Bluecore
Developed a high-performance Vanilla JS engine for Tier-1 retailers (Nike, Reebok). Drove large-scale campaign execution and analytics across North America. Awarded the VIC Award for scaling campaign architecture with zero-latency execution.
Robotics & AI Research Intern · JMoon Technologies
Engineered a CNN-based facial recognition system for an autonomous home robot, optimized for real-time inference on ODROID-XU4 SBCs. Developed a few-shot learning pipeline requiring only 4-5 training images.
Skills
Projects
Audience Builder AI Agent
An AI-powered agent that converts natural language descriptions into runnable audience segments, dramatically reducing manual configuration time.
High-Performance Vanilla JS Engine
A custom JavaScript engine powering campaign execution for Tier-1 retailers like Nike and Reebok across North America.
jQuery to Custom Library Migration
Replaced jQuery with a lightweight 6KB custom library, achieving 5x bundle size reduction and significant execution speed improvements.
Articles
Building a Self-Improving RAG System for AI Agents
Lessons from building a RAG database that improves itself through user feedback loops for an AI-powered audience builder.
Migrating Python 2 to 3 Across 450 Repositories
How we used AI-assisted code analysis to automate and orchestrate one of the largest Python migrations I've ever been part of.
Vanilla JS Engine: Lessons from Building a Zero-Dependency Runtime
What I learned architecting a high-performance JavaScript engine for Fortune 500 e-commerce campaigns.
Education
B.Tech in Computer Science
Hindustan College of Science & Technology, Mathura
2014 — 2018 · 78.78%
Achievements
VIC Award · Bluecore
Peer-nominated award for excellence implementing high-value features in JS integration for major brands like Nike, significantly increasing their revenue.
20202nd Runner Up · E-Yantra Robotics Competition
IIT Bombay. Built a robot from scratch and performed tasks in a pre-defined environment and timeline. Presented at IIT Bombay after reaching the finale.
2016