My Journey from Web Developer to Full Stack AI Engineer

The skills I learned, projects I built, and lessons from 20+ months of professional experience in software development and AI engineering.

Where It Started: The Web Development Foundation

My journey began during my B.Tech in Computer Science at SNS College of Engineering. Like many aspiring developers, I started with the fundamentals — HTML, CSS, JavaScript — and quickly moved into the MERN stack (MongoDB, Express, React, Node.js). The web felt like the most immediate way to build things people could actually use.

My first real professional experience came as a Web Developer Intern at The Reciprocal Solutions, where I spent 10 months building interactive tutorials and real-world projects with Next.js and TypeScript. This was where I learned that clean code, good documentation, and developer experience matter just as much as the features you ship.

The Full Stack Leap: AuxaCode Solutions

Moving to AuxaCode Solutions as a Full Stack Developer was a step up in every way. I was building fintech dashboards with complex data visualizations, integrating REST APIs for financial data processing, and creating reusable UI component libraries. Working with React.js, TypeScript, Redux, and Chart.js on production fintech products taught me what it means to write code that handles real money — reliability isn't optional.

The AI Pivot: Why I Made the Switch

While I loved building web applications, I kept getting pulled toward a question: What if the software could think? The explosion of Generative AI, the practical applications of Computer Vision in industry, and the power of NLP to understand human language at scale — these weren't just academic curiosities anymore. They were becoming the backbone of the next generation of software.

I started by building side projects: an AI Meal Planner using the Gemini API, a heritage artifact recognition app with YOLO and OpenCV. Each project taught me something new about the ML pipeline — data collection, model training, evaluation metrics, deployment challenges.

Key Projects That Shaped My AI Skills

Two projects in particular defined my transition:

1. PPE Real-Time Detection System — Building a computer vision system with YOLOv8 and ByteTrack for industrial safety taught me everything about real-time inference, multi-object tracking, and building full-stack AI systems. I wrote about this in detail in my other blog post.

2. Drug Review Sentiment Analysis — Fine-tuning DeBERTa-v3-base for healthcare NLP showed me the power of transfer learning and the challenges of domain-specific AI. Medical text is complex, ambiguous, and requires models that truly understand context.

What I've Learned Along the Way

1. Full-stack skills amplify AI work — Knowing how to build APIs, databases, and frontends means I can deploy AI models into complete products, not just notebooks.
2. Projects beat courses — Every major skill leap came from building something real, not from watching tutorials.
3. The best engineers are polymaths — Understanding both web architecture and ML pipelines gives you a unique perspective that's rare and valuable.
4. Consistency compounds — 20+ months of showing up every day, writing code, and learning from mistakes has compounded into skills I couldn't have imagined when I started.

What's Next

I'm currently working as an ML Engineer at TFrenzy Private Limited, exploring Agentic AI and intelligent automation. The field is moving incredibly fast, and I'm excited to be building at the intersection of full-stack engineering and AI.

If you're a developer thinking about getting into AI — just start building. Pick a problem you care about, find a dataset, train a model, and deploy it. The best way to learn AI engineering is to engineer AI systems.