One person, every discipline, all of it made with AI.

I started a BCA at Manipal in 2024 because I wanted to understand how computers actually work. One semester in, I was more interested in YouTube. I taught myself editing by cutting football clips, built a channel, and then lost it to a hack.

So I went and got paid to edit instead.

ChatGPT had landed the year before and I'd been pulling it apart since — how these models work, temperature, seed, why one prompt lands and the next one doesn't. When generative video arrived in 2025, I already knew how to talk to them. Kling, Runway, Midjourney. Describing things precisely turned out to be the entire skill.

The advertising half I learned by standing close to it. My elder brother Sibtain is a senior creative head, and I grew up around him and his friends. Everything I know about brands, ads and why a thirty-second film works came from being in that room.

The day starts in WhatsApp — groups first, then email, then whoever I'm working with that day. After that it depends entirely on the job.

Most of it runs the same shape: an LLM to think the problem through and build the prompt, then a production tool to actually make the thing. Freepik for images. Higgsfield for video. Claude Code, Antigravity and Codex when the work is automation, scripting or a real build. Claude and ChatGPT are open constantly — for prompting, and for arguing a problem out loud. Gemini when I need to research something properly.

It isn't a fixed pipeline. I pick the tool after I understand the problem, not before.

Brand films, and AI video of any kind. Product statics and key visuals for social.

I'd rather not narrow it past that. I like doing all of it, and the range is the point — one person covering what normally takes a team.

Though if nobody's assigning me anything, I drift the same direction every time: research, code, and building things on the web.