01Available
Have a conversation with your data.
Express requests through AI chat and explore data through the existing SQL query workflows. Project commands connect conversation to your application.
FROM KNOWLEDGE TO ACTION, IN ONE WORKSPACE
Your business.
Your workspace.
Your control.
Bring your knowledge, screens and workflows together. We’re building OGAN AI into a workspace that reports, creates and acts with your approval.
Existing application platform. Evolving agent experience. Your own AWS server.
OGAN / WORKSPACESample · In development
From context to decisions to action.
Summarize this week’s work, prepare a report view and suggest follow-up steps.
Weekly work overviewView draft
↗Sales conversationsSummary and follow-ups
◫Support recordsReport by topic
↳Follow-up tasksSuggested priorities
Sources and access scope accompany the report.
Plan the requestShow tasks and affected records
Check permissionsPerson, data fields and connection scope
Review and actExplicit approval for code, domains or workflows
Design sample: no tasks or external actions run.
Company knowledgePermitted documents and business records
Authorized external sourcesConnection scope and source date
An answer draft with source references
Design sample: no real documents or model calls.
Your dedicated server8 GB RAM · 150 GB disk
01 / THE FOUNDATION TODAY
Data, screens and code.
One connected foundation.
The existing OGAN AI platform is the foundation of the new workspace. We are evolving its connected capabilities, from chat and data to management screens and application generation.
02Available
Build screens around your work.
Generate data fields, forms, CRUD operations, a Web API and an admin panel from your schema. Manage business data inside your application.
03Available
Create it. Keep the source.
.NET Core generation, source downloads and GitHub pushes in one platform. Software development is one of the things you can do with OGAN AI.
From conversation to commands. We are extending direct model-API chat and the existing command API as a shared foundation for new agent and MCP integrations.
Explore commands02 / THE NEXT OGAN AI
Understands your work. Creates with you.
Acts with your control.
A workspace that starts with a question and can create a report, screen, automation or application. Different needs call for different capabilities, with shared knowledge, permissions and approval.
01
From a request to a workflow.
Agents that plan work, check permissions and act with your approval. Controlled workflows spanning recurring tasks, code, domains and automation.
Agents · automation · approval02
Screens that report and guide.
Reports, dashboards and business screens shaped by your request. Turn data into useful views and findings into follow-up tasks.
Reports · dashboards · screens03
Internal knowledge. External context.
Chat that considers internal documents and authorized external sources together, with referenced answers and visible source freshness.
Company knowledge · external sources04
The right workspace for each person.
Customer 360, role- and person-specific interfaces, and field-level permissions. Different workspaces for sales, service and operations on shared information.
Person · role · data field05
Let your tools work together.
API and MCP connections, code imports from authorized GitHub repositories, and approved actions across business applications. A clear access scope for every connection.
Integrations · API · MCP06
More options for your data.
PostgreSQL and MySQL support, data synchronization and test/live management. A data foundation that builds on the existing schema through controlled transitions.
POSTGRESQL / MYSQL / DATA SYNC03 / EXPLORE THE POSSIBILITIES
One request.
A workflow with context.
Choose a scenario. Explore how information is found, where permissions matter and when your approval is needed.
Sample data · In development
Sales workspace
“Summarize the latest Atlas conversations and draft a proposal.”
STEP 01
The customer story, in one place.
The agent considers the customer record, conversation notes and product catalog you can access.
- Information sources
- Customer record · conversation note · product catalog
- Permissions and control
- Assigned customers only. Cost fields are hidden from this role.
STEP 02
From context to a concrete next step.
A conversation summary and proposal draft are prepared. A missing delivery date is flagged for user input, not assumed.
- Information sources
- Conversation summary · permitted product prices
- Permissions and control
- Drafting is permitted. Discount approval is separately required.
STEP 03
Every external action is visible.
The recipient and content are shown. Sending and record updates are planned only after explicit user approval.
- Information sources
- Proposal draft · recipient details
- Permissions and control
- This example sends no email and makes no database changes.
Support workspace
“Review this support request, suggest a solution and draft a reply.”
STEP 01
Understand the request with its history.
The support record, previous requests and accessible product guide are considered together.
- Information sources
- Support request · product guide · request history
- Permissions and control
- Only records permitted for the support team. Payment details stay hidden.
STEP 02
See what the proposed answer is based on.
The proposed solution includes the relevant guide reference. Uncertain points are flagged for a specialist.
- Information sources
- Product guide · relevant solution note
- Permissions and control
- Reply drafting is permitted. The request is not closed automatically.
STEP 03
You choose the response to the customer.
Edit the reply, route it to a team or approve sending. See which record each action would affect.
- Information sources
- Reply draft · support record
- Permissions and control
- This example neither updates a ticket nor sends a message.
Knowledge workspace
“Prepare a first-week plan for our new teammate.”
STEP 01
Bring scattered information into context.
Permitted onboarding documents, the team guide and role description are brought together. Restricted documents are excluded.
- Information sources
- Onboarding guide · team guide · role description
- Permissions and control
- HR salary records and personal documents are out of scope.
STEP 02
A plan shaped around your company.
First-week tasks are drafted with source references. A source whose freshness is uncertain is explicitly flagged.
- Information sources
- Permitted training materials · team practices
- Permissions and control
- Read access and drafting only. No additional access is granted.
STEP 03
Complete the plan with its owners.
The manager reviews tasks and owners. Adding them to a calendar or task list is designed as a separate approved action.
- Information sources
- First-week plan draft
- Permissions and control
- This example assigns no tasks and creates no calendar entries.
04 / MORE WAYS TO WORK WITH AI
One workspace.
Your choice of AI.
Start inside OGAN AI. We’re developing paths to continue with your own model connection or your preferred agent tool.
Existing foundation
OGAN AI
Start with the platform’s chat.
OGAN AI’s existing chat and command infrastructure uses direct model APIs, primarily DeepSeek. The new experience builds on that foundation.
The samples on this page make no real model calls.
Planned
YOUR OWN API KEY
Choose your model and account.
Connect a supported provider’s API key with project-scoped access, and manage your model choice, usage scope and budget preferences.
API usage is billed separately by the provider.
Planned
YOUR TOOL / SUBSCRIPTION
Continue with your own agent.
Export project context and work with external clients such as Codex through a local bridge. You define the scope of code, tasks and server access.
Your own plan’s access conditions and usage limits apply.
DOWNLOADABLE SAMPLE TEMPLATE
A clear starting point for your next tool.
Explore the approach with AGENTS.md, project context and an SSH configuration template. The pack contains no real project data, server addresses, passwords or keys, and makes no automatic connection.
An external client and an API are different usage paths. About Codex account and API access
05 / YOUR BUSINESS. YOUR FOUNDATION.
Room for your
next business idea.
A foundation to build on.
Business data, screens and source code in a runtime dedicated to your project. We are developing new reporting, integration and agent capabilities on that foundation.
Start your own projectDEDICATED WORKSPACE
Dedicated to
your project.
8GBRAM
150GBDISK SPACE
- AWS
- Business data + application screens + source code
- A shared runtime for new capabilities
Current server configurationGet started
A FEW CLEAR ANSWERS
Good questions.
What exists today, and what is being developed? Here is the distinction.
Read the documentationIs OGAN AI only for software development?
No. Our vision brings conversations with knowledge, automation, reports, dashboards, business screens, integrations and agent management into one workspace. Code generation is one capability within that whole. The status of new features is marked on this page.
What can OGAN AI do today?
The existing platform includes AI chat, SQL query workflows, API commands, schemas and data fields, and form and admin-panel generation. .NET Core generation, source downloads, GitHub pushes and a project-specific server complete this foundation.
Are the new agent features available now?
Not yet. Sections marked “In development” show the product direction and design preview. The scenario explorer uses sample data; it sends no requests to a model and changes no records.
Can I access the generated source code?
Yes. Existing source-code downloads and GitHub push features are preserved. Importing project code from an authorized GitHub repository is on the development roadmap.
Which databases are supported?
The current generation pipeline uses Microsoft SQL Server. PostgreSQL and MySQL support will be developed alongside the schema and code generation system. This page does not automatically migrate an existing project to another database.
Will I be able to use my own model or agent subscription?
Connecting your own API key and continuing through your own account in external agent clients such as Codex are planned options. API usage is billed separately by the provider; external clients use your own plan’s access conditions and limits. The downloadable starter pack is currently a generic template only.
What does a dedicated server provide?
The application and database run together on a project-specific AWS server. The current configuration includes 8 GB RAM and 150 GB disk, providing a shared runtime for business data, application screens and source code.
THE NEXT IDEA IS YOURS.