User Guide
A practical guide for teams and decision-makers - get the most out of the Agentic360 platform
Overview
Welcome to Agentic360 - your AI platform for regulated business
Agentic360 brings specialized AI agents, your own documents and built-in compliance together in one secure workspace - so your teams get reliable, source-backed answers instead of guesswork.
Every answer is grounded in your organization's knowledge and the relevant regulations, and every sensitive action is logged. That makes the platform safe to use even for confidential, regulated work.
This guide explains the platform from a business perspective: what each area does, when to use it, and the value it delivers - no technical background required.
At a glance
Everything connected in one secure workspace
Step by step
- 1Open Chat and ask a real business question
- 2Upload a document and ask the agent about it
- 3Check your status in the Dashboard or Executive Cockpit
Tips
- New here? Open Chat and get a useful answer in under a minute.
- Leaders: jump straight to the Executive Cockpit for a board-level status view.
- Every screen is available in German, English and Spanish - switch language anytime in the top bar.
Chat
Get reliable answers in plain language
Chat is where the daily work happens. Ask a question the way you would ask a colleague, and a specialized agent answers using your documents and the relevant rules.
Answers cite their sources, so you can verify them before you act. Conversations are saved, so you can continue later or hand them over to a colleague.
You can type, attach an image or a document, or simply speak your question - the assistant understands all of them.
When something is clearer as a picture, simply ask for a diagram - for example “show the approval process as a flowchart”. You get a finished, interactive diagram you can enlarge, search and export as an image or PDF.
Results as files: just say what you need - "Turn this into a report" gives you a Word file, "… as PDF" a PDF, "… as a presentation" a PowerPoint file and "… as an Excel spreadsheet" an Excel file. The download link appears right in the answer. Every file is marked as AI-generated, and Excel spreadsheets contain values only, no formulas. Whether this is available depends on how your workspace is set up.
Projects: if you work in a project, select it with "Project" above the conversation list. The assistant then uses only that project's material and general documents without a project. Confidential project material thus stays within the project team - workspace administrators, too, see it only as members. Your conversations are visible to you alone. For another project, start a new conversation.
Confidential details stay protected: if input redaction is switched on, the platform replaces client names, IP addresses, credentials and similar details with placeholders before a language model sees your message. In the answer you normally read your original values again. More in the section "Security & Roles".
Links: web addresses you write into your message yourself stay in the answer. Links the assistant invents are removed by the platform - so you never pass on false addresses.
At a glance
Step by step
- 1Pick the agent that matches your topic
- 2Type, speak or attach your question
- 3Read the answer together with its sources
- 4Refine with a follow-up or export the result
Tips
- Be specific: state the goal, the context and the decision you need to make.
- Attach the relevant document so the answer is based on your data, not generic knowledge.
- Check the cited sources before forwarding an answer to clients or management.
- Ask for a diagram when you need to explain a process or a relationship - the result can be exported or saved as evidence.
- If you need the result as a file, say so explicitly: "Turn this into a report", "Turn this into a presentation" or "Turn this into an Excel spreadsheet".
- Working on a project? Select it before you ask about the project material.
Image Upload
Upload images for visual analysis
Document Upload
Share documents for AI-powered processing
Voice Input
Talk to the assistant using your microphone
Agents
Specialized AI experts for each business area
Agents are ready-made AI experts - one for compliance, one for research, one for document analysis, and so on. Each is tuned for its domain, so you get focused, professional results.
Some agents are available platform-wide; others are configured specifically for your organization and its regulations.
Choosing the agent that matches the job - instead of relying on a single general assistant - makes a clear difference in the quality of the answer.
At a glance
Step by step
- 1Open the agents overview
- 2Browse agents by area of expertise
- 3Open an agent to see what it can do
- 4Start a chat with the right agent
Tips
- Read the short description to see what each agent does best.
- Use domain agents (for example GDPR or MaRisk) for regulated questions - they know the framework.
- Ask your administrator to add a custom agent for a recurring need in your team.
Workspace
Turn your documents into answers
Upload contracts, policies, reports or spreadsheets and let the agents read them for you. Ask questions, request summaries, or compare documents - in seconds instead of hours.
Supported formats include PDF, Word, Excel and more. Files are stored securely, kept separate per organization, and used only for your own analyses.
Organize documents in your own folders under Workspace: give each folder a name and an optional color when you create it, and rename, move or nest it at any time. Access rights are set per folder and cascade to subfolders and files.
Word, Excel and PowerPoint files open and edit directly in the browser, with several people at once when needed. Saving writes back to the same file. The Workspace always keeps the current version; version history is not available yet. Agents answer from the new state afterwards.
Project folders: if a folder or knowledge set belongs to a project, only the project members see the documents in it - including its subfolders.
Retention with reminders: a document can be given a date until which it is kept and a person responsible for it. When the date is reached, that person gets an entry under “Tasks” and - unless they have opted out - an e-mail. In the document view they then decide: "Delete" or "Set new date". Without a decision, a reminder follows after 30 days. The platform never deletes anything on its own, and statutory retention obligations always take precedence: if deletion is not yet allowed, you set a new date.
At a glance
Step by step
- 1Open Workspace and create a named folder via New folder
- 2Click Upload and select a file or a whole folder
- 3Wait for secure processing
- 4Open Office files directly in the browser to edit them when needed
- 5Ask the agent about the document in Chat
Tips
- Use meaningful folder and file names - it speeds up search for everyone.
- Upload the source document before asking, so the answer is grounded in it.
- Make small corrections directly in the browser instead of downloading the file and uploading it again.
- Remove documents you no longer need to keep your workspace tidy and compliant.
- If you receive a retention task, decide promptly - without your decision the document simply stays where it is.
Knowledge & Search
Search across everything your organization knows
The knowledge area connects all your documents, agents and past work into one searchable memory. Ask a single question and get an answer drawn from everything relevant - not just one file.
This is how you stop re-explaining context and re-finding documents. Institutional knowledge stays available even when people change roles or leave.
Every result points back to the underlying source, so the answer stays traceable and trustworthy.
At a glance
Step by step
- 1Open Knowledge
- 2Type your question in everyday language
- 3Review the answer and its sources
- 4Open a source to dig deeper
Tips
- Ask broad questions here; ask narrow ones in Chat with a specific document attached.
- Use it for onboarding - new staff can ask the platform instead of interrupting colleagues.
- The better you curate your documents, the sharper the answers become.
Workflows & Automation
Automate routine work - safely
Workflows let you automate recurring tasks - sorting incoming email, drafting replies, routing documents for approval, generating reports - without writing any code.
You design the flow visually as a series of steps. A built-in test mode runs it safely first, so nothing is sent or changed until you are confident.
Automation frees your team from repetitive work and makes results consistent, faster and fully auditable.
At a glance
Step by step
- 1Open the Workflow Builder
- 2Add the steps you need to the canvas
- 3Run it in simulation mode to check the result
- 4Activate it once it behaves as expected
Tips
- Always run a new workflow in test (simulation) mode before activating it.
- Start with one painful, repetitive task - then expand once it has proven itself.
- Add an approval step for any action that contacts customers or changes records.
Integrations & Connectors
Connect the tools your business already uses
Connect your own Gmail, Outlook, Calendar, Google Drive or OneDrive so agents and workflows can work with your real data - read a mailbox, schedule a meeting, pull a document.
Connections use secure sign-in, are limited to exactly what you approve, and can be disconnected at any time. You stay in control of access.
For technical teams, the platform also speaks the open MCP and A2A standards, so it can connect to other tools and AI systems.
At a glance
Step by step
- 1Open Settings and go to Integrations
- 2Choose a service to connect
- 3Sign in and approve the requested access
- 4Use it in Chat or in a Workflow
Tips
- Connect only the accounts you actually need, and review them from time to time.
- Use connectors to let an agent act for you - for example, summarize today's inbox.
- Disconnect a service in one click when a project ends or a colleague leaves.
Compliance
Meet regulatory requirements with confidence
The compliance area assesses your organization against frameworks such as the EU AI Act, GDPR, MaRisk and many more - and shows you exactly where you stand.
Specialized agents answer regulatory questions with citations, while the compliance scanner identifies gaps and recommends concrete measures.
Everything is documented and exportable, so you walk into an audit prepared instead of scrambling at the last minute.
At a glance
Step by step
- 1Open the Compliance area
- 2Choose the relevant framework
- 3Review the assessment and the gaps
- 4Document your measures and export the report
Tips
- Run a scan to establish a baseline before you plan remediation.
- Keep your compliance evidence current - outdated documents weaken an audit.
- Export reports regularly so you always have an audit-ready snapshot.
Executive Cockpit
A board-level view of where the organization stands
The Executive Cockpit condenses compliance, risk, finance and operations into a single status picture - designed for leadership, not for specialists.
Each indicator shows whether it needs attention. Red and amber items point to the owner and the next action, so meetings shift from 'what is the status?' to 'what do we decide?'.
Snapshots over time show whether things are improving, and one click produces a PDF report for the board.
At a glance
Step by step
- 1Open the Executive Cockpit
- 2Read the overall score across the four domains
- 3Click the items that need attention
- 4Export a board-ready PDF
Tips
- Start each leadership meeting with the cockpit's overall score and trend.
- Click red and amber indicators to see who owns them and what comes next.
- Generate the PDF before board meetings for an instant status report.
Dashboard
Your personal starting point
The dashboard is your home base: recent activity, quick actions and key figures at a glance, tailored to your role.
From here you can jump straight to the area you need and keep an eye on how the platform is being used.
It is the fastest way to pick up where you left off.
At a glance
Tips
- Check your notifications here so nothing important slips through.
- Use quick actions for the tasks you do most often.
- Watch your usage trends to plan capacity and budget.
Security & Roles
Keep access controlled and data protected
Security settings let administrators control who can see and do what. Roles ensure each person has exactly the access they need - no more, no less.
Protect accounts with two-factor authentication, and rely on strict separation between organizations so your data never mixes with anyone else's.
Sensitive actions are recorded, giving you an audit trail for accountability and compliance.
Input redaction - confidential data stays with you: your administrator can switch on that confidential details are replaced with placeholders such as [IP_1] or [CLIENT_2] before a language model sees a message or file. The model works only with the placeholders. In the mode "Mask & restore" the platform puts the original values back into your answer; in the mode "Mask only" the placeholders stay.
What is protected - depending on your administrator's selection: client names from your company's list, IP addresses, host and domain names, web addresses, credentials such as passwords, keys and tokens, cloud account IDs, MAC addresses, usernames, file paths, e-mail addresses and Singapore ID numbers (NRIC/FIN). This applies to your messages, attached files and library documents as well as to web searches, image requests and research the platform runs for you. Text in images is read and passed on redacted; the image itself and voice recordings are not sent to the model while redaction is on.
Good to know: redaction works with fixed detection rules. It recognises people's names only if they are on the client list. Occasionally a harmless technical detail is replaced as well - in the answer it still appears in its original form as long as the mode "Mask & restore" applies.
Traceability: chats, uploading, viewing and downloading documents, and changes to permissions are logged; chat logs are kept for one year by default. You see your own entries under Settings → "My activity". Administrators and auditors see the log of the whole workspace, with confidential details redacted. The unredacted view requires a reason, and that access is recorded too.
Security work in chat: for your protection, the platform blocks messages containing program code, SQL or scripts by default. If your team works on threat models or security tests, your administrator can set such patterns to "Warn" - the message then goes through and the match is logged. Attempts to get around the assistant's safety rules always stay blocked.
At a glance
Input redaction at a glance
Step by step
- 1Open Settings and go to Security
- 2Enable two-factor authentication
- 3Assign the right role to each team member
- 4Review the audit log periodically
Tips
- Enable two-factor authentication for every account, especially administrators.
- Review roles and permissions regularly, and remove access when people leave.
- Use the audit log to investigate anything unexpected.
- Working with client data? Ask your administrator whether redaction is switched on and whether your clients are on the list. Administrators maintain both in the admin console under Compliance & Governance → "Input Redaction".
- Administrators: use the unredacted log view only for a concrete reason - your reason is stored.
Settings
Make the platform yours
Settings is where you manage your profile, language, notifications and connected services.
Administrators also manage organization-wide preferences, branding and integrations from here.
A few minutes here makes daily work smoother for you and your whole team.
Under "My activity" you see what is logged about your chats. Credentials, technical identifiers, e-mail addresses and client names from the list are always hidden there; people's names are not. Administrators and auditors find the log of the whole workspace in this place.
Step by step
- 1Open the user menu
- 2Select Settings
- 3Adjust the section you need
- 4Save your changes
Tips
- Keep your contact details and notification preferences up to date.
- Set your preferred language - the whole platform follows it.
- Admins: configure integrations and branding once for everyone.
- Look into "My activity" now and then to see what is recorded about your chats.
FAQ
Frequently asked questions
Here you will find answers to the questions we hear most often.
If your question is not covered, please contact our support team - we are happy to help.
We update this page regularly as the platform evolves.
Why do I see placeholders such as [IP_1] in an answer? - Redaction in your workspace is set to "Mask only", so the placeholders stay visible. If you need the original values in answers, talk to your administrator.
Why was my message with program code blocked? - Code, SQL and script patterns are blocked by default for your protection. If you need such examples for security work, your administrator can set them to "Warn".
Why is a link the assistant mentioned missing? - Links the assistant invents are removed, so no made-up addresses are passed on. Web addresses (https://…) that you write into your message stay in the answer.
Why can't I see a document in chat? - If it belongs to a project, it is available only to that project's active members. Ask to be added to the project and then select it in chat.
Who can see my chats in the log? - You yourself under "My activity", and your company's administrators and auditors in the workspace log - redacted by default, and unredacted only with a recorded reason.
Tips
- Use the search box at the top left for a quick answer.
- Read related sections of this guide for more context.
- Contact support for anything complex or account-specific.