| Years | Role | Scope |
| 2026 | Software Engineer | Full-Stack Engineering, System Design, AI Pipelines, Video & Image Generation, Automation, Growth |
Short-form content that actually performs is slow and manual: research what's working, script it, generate visuals, edit, and post to many accounts every single day. By hand it doesn't scale, so I wanted a machine that runs the whole loop.
It studies viral formats, generates original scripts, produces slides and short-form video, auto-edits them into finished reels, and distributes them across social media accounts on a schedule. What began as a semi-manual workflow is becoming fully autonomous, and the same engine is the foundation for a SaaS that lets anyone run their own content machine.
A content machine, not a content calendar
Manual short-form doesn't scale: by the time you research a trend, script it, design the slides, edit the video, and post it across accounts, the window has closed. The bigger idea is variation at scale. When one slide format works for a specific segment, the engine spins it into many styles automatically and reuses it to reach completely different audiences. Spurt automates that entire loop and distributes across secured channels and platforms, and the end goal is an agent that runs it and improves a little at every step, day by day.
From viral format to original script
Some edits rack up millions of views, so the engine breaks them down: the editing, the audio, the context, the imagery, and what actually lands with the target audience. An LLM then writes an original script in the right voice and structure, so every piece starts from proven patterns instead of a blank page, and the whole analysis runs on an easy loop.
Slides and video, generated
From the script, Spurt generates carousel slides and short-form video automatically: visuals, layout, and pacing produced end to end without manual design. The same template spins into variations for different segments, a dynamic content creation system that runs on its own.
Learn from what performs
After publishing, Spurt tracks videos across social platforms so the loop doesn't stop at generation. Views, engagement, and momentum signals feed back into what formats the system should study next.
I own Spurt end to end: the system design, the generation pipeline (scripting, slides, video, and auto-editing), and the distribution layer that publishes across accounts. It's still being built and under active development.
It's still in development, with a few templates already running in production. Along the way I learned how to automate video editing and produce a high volume of carousels from a semi-automatic system.
Distribution is the hard part, not generation. Posting content to many accounts on a schedule, reliably, is what turns a generator into a real growth machine.
Let's build
something
together.