RPA & automation
Hand off repetitive work
Opening a portal, collecting information, filling the same fields, moving files… Turn everyday manual tasks into a clearly defined workflow.
AI AND AUTOMATION FOR YOUR BUSINESS
Less repetitive work.
More room to grow.
Collecting information, preparing reports, entering the same data again… Describe the work that takes up your day. Start with your data, screens and AI chat in your own workspace, and define the automation your business needs.
Start in the language of your business. Explore technical details when you need them.
Your business does not have to fit a fixed template.
- Your data
- Screens for your work
- Actions under your control
- Your own AWS environment
MORE THAN A SOFTWARE TASK
Start with the work that takes up your day.
Let RPA, automation and AI serve the same purpose: less repetition, clearer information and work with a clear owner. Use these areas to describe your needs.
Data collection & updates
Keep information coming
Define what to collect from permitted websites, business apps and files, and how often. Track changing data and missing sources together.
Publishing & sharing
Get results to the right place
Define who receives a report, product update or content, where it goes and when. Review information before it leaves your workspace.
AI work assistants · Agentic AI
Move from questions to next steps
More than a chat response: understand the task, bring information together, propose next steps and connect them to permitted actions.
Reports & dashboards
See what matters at a glance
Sales, service, inventory or operations. Bring the information, follow-up work and screens your team needs into one workspace.
Integrations & business screens
Make your tools work together
Connect your applications instead of entering the same data repeatedly. Shape your screens around your business, not a fixed template.
Different tasks need different setups. Define sources, permissions and the intended result together. Features available in your account
MEASURE. LEARN. IMPROVE.
Work that improves, not just runs.
Start with “Where are we losing time?” or “Which campaign is underperforming?” Describe your question in the analysis workspace, select your data and use measurable findings to choose the next step.
Process and data mining
Examine the steps an order, proposal or support case passes through. See long transitions and repetitions; check missing fields, exact duplicate records and numeric distributions. Turn a table into a clear picture of your work.
Measure digital marketing
“Compare campaign spend with conversions.” See spend, cost per conversion and revenue-to-spend ratios for comparable periods. Spot measurement gaps; do not base a budget decision on a single number.
Connect findings to improvement
Choose a finding to review, describe the change you want and save the request with a success measure. After a change, measure again using the same definitions: did it really improve?
The analysis workspace is not built around CSV uploads: it takes an authorized snapshot from your project database and runs Python data quality, process mining, marketing measurement, forecasting and classification flows. Process analysis needs event history, not only a current table. Results can feed a saved automation, report or approved improvement request. Data is not sent to a model provider.
How do prompts connect marketing, integrations and continuous improvement?
“Compare campaigns every week; tell me which have no recorded conversions and suggest a review workflow.” Such a request defines the source account, data period, schedule, target app and approval boundary together. Preparing content, updating a CRM and publishing require separate permissions.
Our continuous-improvement approach: measurement → finding → proposal → authorized approval → application → re-measurement. Analysis and request preparation are available here; continuous collection, automatic fixes in connected apps, ad management and publishing are not activated by this screen. Those steps need project connections, scoped approval and failure/recovery controls.
A STARTING POINT FROM YOUR OWN WORK
Describe the result you want.
Choose a use case to see where to start. Explore the source, routine and review point in a straightforward workflow.
YOUR REQUEST
Check supplier information every morning and report price changes.
What information do we start with?
Permitted supplier pages, files or business apps. Include only the fields and access permissions needed.
When and under which rules will work proceed?
Choose the days, time and time zone. Define what happens when a source is unavailable and who is notified.
Where does the result go, and who approves it?
A price-change report, freshness information and records to review. Set separate approval boundaries for updates and notifications.
YOUR REQUEST
Bring new product information together and prepare content for the website and team.
What information do we start with?
Define sources for product records, descriptions and images. Make missing fields visible before publishing.
When and under which rules will work proceed?
Separate content and recipients for a website, business app or team notification. Choose the publication time and reviewer.
Where does the result go, and who approves it?
A publishing workflow with a clear destination and approval. Include duplicate prevention, result tracking and recovery needs.
YOUR REQUEST
Summarize customer conversations and prepare a proposal and follow-ups.
What information do we start with?
Bring together customer records, notes and product information you can access. Team members should see only permitted information.
When and under which rules will work proceed?
Define proposal fields, follow-up dates and owners. Do not assume decisions such as discounts or delivery dates.
Where does the result go, and who approves it?
Proposals and follow-ups connected to customer context. Review the content, recipient and messaging permission separately.
YOUR REQUEST
Review an incoming request, find the right information and prepare a reply.
What information do we start with?
Use request history and permitted product guides together. Keep personal information and restricted records out of scope.
When and under which rules will work proceed?
Show which document supports the answer. Define when incomplete or sensitive requests should go to a specialist.
Where does the result go, and who approves it?
A sourced reply and follow-up work for the right team. Separate permissions to review, send and close a request.
YOUR REQUEST
Prepare an answer using our internal documents and permitted external sources.
What information do we start with?
Which team documents and external sources may be used? Decide separately where confidential information may be processed.
When and under which rules will work proceed?
Show source links and freshness together. Flag points that need human review when information is old or contradictory.
Where does the result go, and who approves it?
Turn information into a report, team guide or follow-up task. Keep sharing permission separate from permission to read.
AUTOMATION WITH CONTROL
Delegate the work. Keep the decisions.
Reading information, changing it and sharing it are different permissions. Keep those boundaries visible when defining a workflow.
- Access for each person
Define who may see each record, field and screen.
- Automatic and approved steps
Set separate boundaries for messages, publishing, important changes and spending.
- Follow the outcome
Make failures, retries and escalation part of the workflow.
SCREENS THAT FIT YOUR WORK
Your fields. Your way of working.
A customer record, request form, order list or record summary. Define your fields, design the screens and consider how different teams should see the same information.
Design your screensInstead of disconnected tools
Keep data, screens and follow-up work in the same project context. Technical teams can continue with code generation, source downloads and existing project commands when needed.
Details for your technical teamROOM FOR YOUR BUSINESS TO GROW
Your own environment. Capacity that can grow.
Your application and database run together in an AWS environment dedicated to your project. Starting resources can be expanded as your needs grow; your business does not have to fit one server size.
Capacity increases are planned around users, data volume, workloads and cost.
STARTING CONFIGURATION
A starting point, not a ceiling.
- Memory
- 8 GB
- Storage
- 150 GB
AWS resources can be increased for more data or heavier workloads. The right configuration and migration steps are agreed separately.
Start with what your business needsMAKE YOUR AI CHOICES CLEAR
A path that fits your work and data.
You do not need to memorize model names. Understand where your data is processed, what access is allowed and how costs work.
Start with OGAN chat
Use existing chat and project commands to explore data and describe what you need. Data sent to a model provider is a different boundary from data on your project server.
Define your preference
Save provider, environment, language and purpose preferences in Project Settings. A saved preference alone does not activate a connection or change the current chat model.
Consider your own account
Your own API account, a tool subscription and a privately hosted model have different access and cost conditions. A subscription does not grant unlimited use of every model API.
For your technical team: source code, connections and starter documents
You do not need these files to use OGAN. This section is for people working with a developer or an external coding tool.
What are the downloaded .md files for?
They are readable text documents in Markdown format, not an executable program. AGENTS.md describes working rules, PROJECT_CONTEXT.md provides project context, and README.md explains how to use the template.
The SSH file in the ZIP is also only a configuration example. The kit contains no real project, server address, password or key. It does not connect to a server, create an account or install automation.
Download technical starter documents (.zip)The existing technical foundation
The foundation includes .NET Core generation, SQL Server, PostgreSQL and MySQL runtime support, schema-based fields, API commands, source downloads and GitHub pushes. Project connections and provider selection are parameterized in generation; external connections such as MCP and private-model pairing are added with their own permissions and approvals.
The kit does not grant permission. Share only authorized project information with external tools, and do not overwrite your existing files.
Technical documentationYOUR QUESTIONS
Straightforward answers.
Understand workflows, permissions, capacity and what your account includes.
Do I need technical knowledge?
Start by describing your work, the information you use and the result you need. Continue with data fields, forms and business screens. Technical requirements for custom connections and server work are defined separately; not everyone needs to learn code or commands.
What is the difference between RPA, automation and an AI assistant?
RPA handles repeated screen interactions such as clicks and data entry. Automation connects rules to a schedule or event. An AI work assistant can interpret information and propose a next step. They can work together, with human approval for sensitive actions.
Which features are available in my account today?
AI chat, data queries, person and field permissions, forms/admin screens, reports and analysis workspace, process mining, data quality, forecasting, classification, rerunning saved analyses and approved automation drafts are available in the same project context. Real RPA, scheduled collection, publishing or record changes in connected applications require their connection, permission and approval package to be enabled separately; a use-case description alone never starts an action.
Will actions run without my approval?
Writing a request or exploring a use case does not grant permission to act. Define which steps may be automatic and which need human approval. Sending messages, publishing, changing records and spending limits are considered separately.
Is server capacity limited to the starting configuration?
No. 8 GB of memory and 150 GB of storage are a starting configuration, not a ceiling. AWS resources can be increased as users, data and workloads grow. Capacity, cost and migration needs are reviewed together; upgrades are not assumed to be automatic or interruption-free for every workload.
What happens to my data and generated code?
Your application and database are managed in an environment dedicated to the project. Authorized downloads and GitHub pushes give access to the generated source. Data sent to an AI provider must be considered separately: a project-specific server does not mean every model request is processed there.
START WITH ONE TASK
Think of one task you repeat every day.
What information comes in, who does what, and where does the result go? Shape your workspace around that need.
DATA AND ANALYSIS FOUNDATION
From your database to insight, from insight to action.
Project data is not converted into a file first. An authorized SQL Server, PostgreSQL or MySQL connection is read directly for data quality, process mining, marketing measurement, forecasting and classification. Results can feed a report, dashboard or approved scheduled automation. The database choice, connection profile and field permissions stay with the project; private model connections require separate approval.