Ipek is an IDE for automating repetitive data tasks. Users can talk to their own AI and automate: combining APIs, extracting from PDFs and spreadsheets, cleaning up exports, and more. The visual canvas hosts the deterministic workers that users can execute without per-task fees.
Workers can fetch from or post to APIs, process local files, transform, merge or de-duplicate data, generate structured outputs, and use configured AI or OCR services when needed. Workflows run on the user's own computer without per-task fees. Every worker's outputs, including API responses, are saved as files in inspectable run folders, making workflows easier to test, review, revise, rerun and schedule.
I built Ipek out of pure frustration with existing cloud automation platforms. Every time I needed to process a large batch of spreadsheets, extract text from hundreds of PDFs, or string together a few APIs, I was hit with ridiculous per-task fees. It felt like I was being punished for scaling my own work. On top of the high costs, debugging those cloud workflows was always a headache. If a run failed, I had to dig through clunky web logs just to figure out what went wrong. I wanted something that ran locally on my own computer. I wanted the freedom to use my own AI models without passing my data through a middleman. Most importantly, I wanted a system where every single step and API response was saved as a normal file in a regular folder, so I could just open it and inspect it. I could not find a single tool that gave me that level of control, transparency, and freedom from subscription traps. That is exactly why I decided to build Ipek myself.