How it works
From setup to a working platform in three steps.
1. Create a workspace and upload your content
A workspace is your project, your client, or your department. Inside, create folders for each workstream — training-videos, client-deliverables, project-website, process-models.
Upload files and they're processed automatically. Videos become streaming-ready players. Images are optimized to WebP with thumbnails. Documents render inline.
2. Turn folders into purpose-built apps
Assign a type to any folder and it becomes a specialized application:
Media gallery — Searchable archive with streaming playback and tagging.
Training course — Structured learning environment with presentations and video modules.
Process modeler — Interactive BPMN editor and simulator.
Website — Multi-language site builder with AI-assisted content.
No configuration needed. The folder is the application.
3. Work together with AI agents
Open any folder and compatible AI agents appear. A content writer for your website. A course planner for training materials. A process analyst for your BPMN diagrams.
Give an instruction or pick a quick action. The agent reads your files, understands the structure, and creates or refines content. You approve, adjust, or let it run autonomously.
Your API keys. Your choice of model — Anthropic, OpenAI, or local via Ollama. Everything stays on your server.
Share with anyone — no accounts required
Generate an access code for any folder. Send the link. Your audience opens it and sees everything — videos, documents, training courses, process models.
No sign-up forms. No password resets. Create different codes per audience. Set expiry dates. Revoke when the project ends.
Three ways to deliver when you're done
Keep it on your platform: Clients access via codes on your infrastructure. You control the environment, they consume the content.
Federate across servers: Connect multiple Aipokit instances to share catalogs across your organization or with partners.
Hand over standalone: Export a standalone binary for the client to run on their own server — their data, their branding, their infrastructure.
What people use it for
Two examples from real workflows.
Running a workshop
You prepare a training session. Your AI agent drafts the agenda from your notes. Presentations run directly from the platform. Session recordings upload and auto-transcode to streaming afterward.
Share one access code with all participants. They revisit the materials weeks later — no accounts needed.
Delivering a client project
The workspace holds process models, interview recordings, training materials, and a project website — all in one place. AI agents analyze processes and draft recommendations.
When the client asks where their data lives, you point to their server. No third-party routing.
Questions? Let's talk.
Aipokit is in active development. Early access for businesses, educators, consultants, and teams who want a platform they own.
Get in touch via aipokit.com or reach out on LinkedIn.