Another happy customer here. Since day one Trigger.dev is powering our AI Automation platform (https://turbotable.ai). It helped us to scale effortlessly and stay on TypeScript stack, easily integrates with tools we use. And the pricing is even cheaper than self-hosting on VPS. Thanks guys!
Thanks for sharing so much detail. I am the CTO & Co-Founder of https://turbotable.ai (landing is outdated, will be updated soon), similar product in the space, but mainly focused on more general automation and data analysis for non-technical teams. OCR is one of the tools in our arsenal and our bet is that LLMs will get better at it. 2 limitations with this approach I can see:
- No reliable grounding, bounding box (for now)
- Context length (we have a solution for this, similar to Zerox by Omni)
Even if in the long run foundation models will not solve OCR completely and reliably, we still have option to develop custom solutions or to integrate with mature players.
From the demo it looks great. And the UX is pretty sick! Can it be used via API in the self-hosted version? If yes, what are the limitations over UI? In my use-case I would like to use it as a RAG infra and together with tool calling and generative UI in the client code (already implemented) it would complete the picture.
As a team we use
- NextJS (AppRouter) for the UI, TailwindCSS + shadcn ui
- Custom CMS built on top of Strapi for the backend
- We use self-hosted GitLab and for every project we have CI/CD by default.
- For every project (even simple landings) we use server from Contabo. It’s like 6$ for 6GB RAM and 400GB SSD
- Everything is dockerized and gets deployed using gitlab-runner.
- Traefik is used as reverse-proxy
- Minio is used as a storage (with some tweaks for caching)
- Postgres is DB of choice
Everything lives as scripts. Only Next UI is developed from scratch for each customer.
