Cross-industry
AI assistant on your own website
An agent on Hawk IVO Force answers questions about your content, points to relevant pages and only starts when visitors explicitly want it to.
Hawk IVO Force
A workforce that gets work done.
Finish more work. Build digital workers that turn assignments into steps, connect tasks and work toward results. Assistance remains a deliberate operating choice. Measure completed work against controlled effort.
Plan work and follow execution
Visualize projects, tasks and dependencies. Plan work packages across linked projects and assign agents. Execution and progress stay in the shared work context.
Configure the platform with an agent
Use an agent to configure and evolve agents, skills, projects and operating rules. You provide goals, data connections and approvals. Setup follows the platform’s existing permissions.
Scale work economically
Activity-based billing ties consumption to use. Model choice, reusable skills and run budgets help control effort and consumption. The agreed plan determines individual billing components.
An operating system for agents
You decide how independently each agent works.
One operating framework supports assistance, human approvals and autonomous work. Choose the permitted scope for each agent and task.
Skills, knowledge and operations in one platform.
Connected skills, integrated knowledge retrieval and cross-provider model access share permissions, rate limits and run budgets. Isolated work environments and traceable model calls belong to the same operating framework. Build capabilities that further agents can reuse.
Assistance
The agent researches and prepares. A person determines the next step.
With approval
The agent prepares an action and executes it after the required confirmation.
Autonomous
An event or schedule starts the work. The agent acts within its permissions and limits without ongoing chat instructions.
One budget baseline for every agent.
Company defaults and upper limits also apply to newly added agents. Individual limits and authorized exceptions remain controlled. If the available budget is insufficient, the next model call is blocked.
Operating model, not a product screenshot. Autonomy does not expand permissions. Executable actions depend on configured connections and approvals.
The foundation
Build once. Use across agents.
Maintain a capability in one place. Other skills reference it and multiple agents use the connected knowledge, so proven practices need not be rewritten for each agent.
Shared base skill
Write to your guidelines
These skills reference the base
- Write emails
- Create presentations
- Prepare contract drafts
Multiple agents use the skills
- Service agent
- Sales agent
- Project agent
Versions · approved sources · permissions per agent and tenant
From knowledge to work steps
Information alone doesn't complete a task.
Hawk IVO Force connects knowledge and work context with reusable work instructions. This lets teams handle tasks in a structured way and provide results for the next step.
Human teams, agent teams and mixed teams work on ongoing operations as well as on change initiatives.
Assistant
Who can cover the early shift on line 2 tomorrow?
Checking qualificationsQualifications checked
Checking availabilityAvailability checked
Applying working-time rulesWorking-time rules applied
AssistantTwo suitable options: a floater from the pool or a shift swap with line 1.
Approved
Write a message …
Behind it: what the agent uses
Know-howQualification matrix
SkillsCheck shift rules
RightsPropose, not approve
Shift leadapproves
Why the foundation matters
Every use adds reusable capability.
Connect skills instead of copying
Maintain writing rules, checks and business practices once; reference them from other skills. Assign skills or let agents discover approved capabilities in the catalog.
Control how work happens
Combine fixed workflows, dynamic tasks and human approvals. Configure tools, access and each agent’s responsibility.
Make agents available where work happens
Choose supported models and tools for the task. Deliveries make agents available through website chat or an embedded interface, with configured access and permitted actions.
Control costs and usage
Set budgets and invocation limits for agents and runs and track usage. Verify limit behavior in your concrete use case.
Possible manufacturing scenario
An absence should not stop the whole order.
A qualified worker is absent. An agent uses approved qualification, availability and order data, applies your skills to evaluate alternatives and prepares a feasible proposal. The shift lead decides.
Following approval, configured actions run and the people involved are informed. The outcome is a confirmed change with a traceable decision. This scenario needs connected planning systems and verified rules; it is not a documented Force production case.
- With approval: The agent prepares an action and executes it after the required confirmation.
- Your invocation limits: Limit how often triggers or published API access points may invoke an agent.
- Budgets per run: Run, daily, weekly and monthly budgets apply together. Expected costs are reserved before model calls.
- Permitted subagents: Choose allowed delegation targets. An agent cannot engage arbitrary other agents.
Control
People make the decisions where their responsibility is required.
Agents act only within the limits of what you approve.
Feedback
New requestfor week 42
Planning agentcreates a draftIn progress
Expert reviews
Shift lead approves
Week 42 planeffective MondayPublished
Log · every decision traceable
10:02Planning agent creates a draft
10:04Expert: line 2 is understaffed on Tuesday
10:06Planning agent revises the draft
10:07Expert has reviewed
10:09Shift lead has approved
Access and approvals
- Agents use only configured connections and permissions.
- Actions in connected systems are carried out in a controlled manner.
Skills and knowledge
- Reusable work instructions as skills.
- Company knowledge from approved sources.
Setup and operation
- We bring software and the know-how to build and scale agent systems, and we help with the setup.
- Costs, capacity and scaling are assessed for each use case.
Use cases
Use cases with Hawk IVO Force.
Hawk IVO Force can be used across industries. These use cases are documented; we are continually collecting more.
For software vendors
Agents in your own software
Software vendors bring agent functions into their product as a white label: with Hawk IVO Force, without building their own agent platform.
- Hawk IVO Force
In practice
How we use Hawk IVO Force ourselves.
The AI assistant on this website is an agent on Hawk IVO Force. You can try it yourself at the bottom right.
In-house use · Software
Hawk IVO: using our own products every day
We use Hawk IVO Scheduler, Hawk IVO Force and Hawk IVO Plan within our own company. One publicly visible example is this website’s AI assistant on Hawk IVO Force. It only loads after the chat is explicitly started.
- Hawk IVO Force
Hawk IVO Plan
Hawk IVO Scheduler
Introducing agents with approvals
How to start with a clearly defined task and where people make the decisions.
Read the articleISO/IEC 27001:2022
Our information security management system is certified for the development and sales of our SaaS solutions for staff and production planning.
Security and ISO 27001Subject-matter responsibility
Knowledge articles and case studies name the person responsible for their content. You can find the profiles on our About page.
The people behind Hawk IVO
Decision guide
How does the approach compare?
Force combines deliberately chosen autonomy with concrete run and budget rules. The clearest examples are per-run budgets with cost reservation before the next model call, and defaults and upper limits for new agents. This comparison separates those controls from infrastructure quotas and usage reporting.
Comparison for users
Microsoft, SAP, n8n and building in-house: from individual agents to business capabilities.
Questions
Frequently asked questions about Hawk IVO Force
How do operations differ from Copilot Studio?
Force focuses on your agent system and the rules for every run: triggers, access, invocation limits, run budgets, allowed delegation targets and workspaces. Budget defaults and ceilings also apply to new agents. Copilot Studio also supports autonomous agents; Microsoft documents monthly limits among its controls. See the operating comparison for the specific differences.
Why not build it ourselves?
You can build a first agent. Further agents add skill versions, permissions, user contexts, model connections, logs and operations. Force provides platform foundations; your business logic and system integration remain yours. Compare the full lifecycle.
How quickly can we start?
With prepared skills and configured access, a first agent can be configured in minutes. A dependable process also needs verified data, permissions and acceptance criteria. Start with one task and show what the next agent can reuse.
Are we tied to one model or cloud?
Force is designed for different model providers, tools and runtimes. Choose supported options for the task; deployment, data paths and availability are agreed for your environment. Not every provider or infrastructure is supported.
Do we need Hawk IVO Plan for this?
No. Hawk IVO Force can be used on its own. Hawk IVO Plan complements it when planning logic or operational interfaces are needed.
How can this grow into a landscape of thousands of agents?
Large agent landscapes need rules that grow with them. Shared policies and controlled execution per run matter. Concurrent capacity is sized against your infrastructure and workload.
Company defaults and upper limits also apply to newly added agents. Individual limits and authorized exceptions remain controlled. If the available budget is insufficient, the next model call is blocked.
Conversation
Start with a clearly defined task.
In a conversation, we clarify which use case suits your team and how you can assess the benefit.
