There is a massive difference between writing a script that works on your local machine and building a system that real people use. As I navigate my BTech in Electronics and Biomedical Engineering, my focus has increasingly shifted from purely academic research to the entrepreneurial side of engineering.

I want to build things that scale. Whether it is an online AI service handling thousands of requests or a physical piece of wearable tech decoding biological signals, the goal is to create products that bridge the gap between complex engineering and everyday utility. Here is a look at how I am approaching this process.

Grasping the Hardware Layer

You cannot build revolutionary wearable tech without a deep understanding of bare-metal hardware. A lot of modern development abstracts the hardware away, but I believe the best neurotech products will be built by those who understand the circuits.

Recently, I built a Bioelectrical Impedance Analysis (BIA) waveform generator using an 8051 microcontroller. The system was designed specifically for tissue impedance measurement. Writing 8051 assembly to interface directly with biological tissue is a gritty, exacting process. You are forced to think about memory registers, clock cycles, and signal noise.

This low-level hardware experience is the foundation I am using to explore the wearable tech space. If we want to build consumer-ready Brain-Computer Interfaces (BCIs) in the future, we have to master the physical sensors and microcontrollers that make them possible.

The Reality of Scaling Software

On the flip side of physical hardware is the challenge of software scalability. To test my ability to deploy and scale an application, I recently launched a Tic-Tac-Toe AI bot on Telegram powered by the Minimax algorithm.

The growth was explosive: the bot acquired over 1,000 users within its first 48 hours. But with rapid growth comes real-world lessons. For example, I quickly discovered a bug in my scraper logic where the system was entirely failing to save user IDs to the database!

It was a brilliant learning moment. Building a startup is not just about writing the core AI algorithm; it is about database architecture, server monitoring, and handling edge cases when hundreds of people interact with your code simultaneously.

Hacking the Physical World

I am constantly looking for ways to merge these two worlds: scalable AI and physical hardware. Participating in the Physical AI Hackathon at TinkerSpace Kochi was a great environment to experiment with this exact intersection. Hackathons force you out of the perfectionist mindset and into the "hacker" mindset: building minimum viable products (MVPs) rapidly, breaking things, and iterating.

The Road to Seyarkai

All of these projects, the 8051 assembly code, the database scaling bugs, the local Linux GPU optimizations, are feeding into my current main focus: my AI startup project, Seyarkai.

I am currently exploring ideas and business models to officially launch Seyarkai. The vision is to build a company that develops both high-utility online AI services and eventually transitions into physical, wearable neurotechnology.

The journey from a breadboard and a local Python script to a fully-fledged startup is long, but documenting the process, the bugs, the hardware revisions, and the scaling challenges is half the fun. Welcome to the journey.