Otorecrutor
The company I work for. I'm the engineer behind the application — the Python scraper API everything feeds on, the agent that runs HR, and the pipelines between them. A real office uses it every day.
This is not a side project. I'm employed as an engineer at the company, and the application's code is mine end to end: what gets built, how it gets built, and what is allowed to run in production.
At the centre of it is a scraper API I wrote in Python. Everything the system knows arrives through that layer, which makes it the part that is not allowed to fail — the rest of the product is only ever as reliable as its intake.
On top of that sits an agent-style LLM core. It tracks everyone's working hours on its own, answers HR questions in plain language, and carries the repetitive post-screen-schedule loop of recruiting. If attendance looks off, it gets flagged.
- Wired the agent into the tools the office already had open all day.
- Shipping a separate HR panel and asking the team to learn a new interface.
- Employed engineer, not a side project
- A Python scraper API I wrote, feeding the whole system
- Working-hour tracking, fully automatic
- In daily use at a real company