Team member
Prepare the renewal briefing for Acme before the client call.
ChannelGate A Makeitfuture product
ChannelGate, the AI operating system we built and use
Give your team one governed place to work with AI agents inside the channels they already use. ChannelGate brings company skills, permissions, memory and human approvals into each request.
Product information and all enquiries continue on channelgate.dev.
ChannelGate
Illustrative workflow ChannelGate
- Identity verified
- Approved skills loaded
- HubSpot and ClickUp checked
Briefing ready with open decisions and source links.
Human approval required before the CRM is updated
Why we built it
AI agents need a company operating layer
A useful agent needs access to company systems, a clear identity and limits on what it can change. As more people and agents join the workflow, separate chat windows and personal prompts become difficult to manage. We built ChannelGate so the company can define the rules once and make them available where the team works.
-
Shared company capability
Approved skills and operating instructions belong to the company, so the quality of the work does not depend on one person keeping the right prompt.
-
Permission-aware access
Each request starts with an identified user. The agent receives the tools and credentials approved for that person and that job.
-
Human control
Sensitive actions pause for approval before the agent changes a record, sends a message or completes another high-impact step.
-
A record of the work
Requests, tool use, approvals and results stay attached to the work, which gives teams a clearer way to review and improve how agents operate.
How it fits
From a team request to governed work
ChannelGate sits between the conversation, the AI agent and the systems that hold company data. Your team can ask for work in a familiar channel while the company keeps control of access and execution.
Where your team asks
- Slack
- Microsoft Teams
- Google Chat
ChannelGate What ChannelGate governs
- Identity and access
- Skills and memory
- Approvals and history
Where the work happens
- CRM and ERP
- Email and documents
- APIs and databases
Inside Makeitfuture
How we use ChannelGate in our own work
We use ChannelGate as the shared entry point for AI agent work across the company. The workflows below show the type of work we place behind permissions, reusable skills and approvals.
-
Sales
Build account context before a call
An agent gathers the approved CRM record, recent communication and delivery context, then returns a briefing in the team channel.
ResultThe owner reviews one sourced briefing instead of searching across systems.
-
Delivery
Turn project context into the next action
The agent reads the permitted project sources, finds open decisions and prepares the next task or handover for a person to approve.
ResultThe team keeps momentum without giving the agent unrestricted access.
-
Operations
Investigate a failed workflow
A specialist agent checks the available run data and system context, then explains where the workflow broke and what should happen next.
ResultDiagnosis is faster, while production changes still wait for a person.
-
Knowledge
Reuse company methods consistently
Shared skills give agents the current company process for recurring work, from research and analysis to preparing a client deliverable.
ResultThe method is maintained once and reused across the team.
Product capabilities
What ChannelGate controls for the company
The product site carries the technical detail. These are the parts that matter when you move from individual AI use to shared company operations.
- 01
Company identity
Connect each request to a verified person and apply the access rules attached to that identity.
See governance and security - 02
Approved connections
Give agents access to the systems and credentials approved for their work, without placing personal secrets in prompts.
See credential controls - 03
Reusable skills
Package company instructions, methods and checks as skills that can be maintained and reused across agents.
See how skills work - 04
Memory and history
Keep useful context and a record of agent work so teams can review results and continue from previous decisions.
See memory and history - 05
Human approvals
Pause sensitive actions for a person to review before the agent writes to a system or contacts someone outside the company.
Explore the product - 06
Deployment choice
Choose the deployment model that fits your security, data and infrastructure requirements.
See deployment options
Choose the right layer
Automation, AI agent or AI operating system?
These layers solve different problems and often work together. The distinction becomes important when several people and agents need governed access to company systems.
| Aspect | Workflow automation | AI agent | AI operating system |
|---|---|---|---|
| Main job | Runs a defined sequence of steps | Pursues a goal and chooses actions | Coordinates people, agents and company controls |
| Best fit | Predictable processes | Work that needs judgement | Shared agent use across a company |
| Access model | Fixed workflow credentials | Tools configured for one agent | Identity and policy-based access |
| Human control | Exceptions follow workflow rules | Approval depends on its design | Approval gates are part of the operating layer |
| Company memory | Stored data and workflow state | Context for its assigned work | Shared skills, context and work history |
Built by practitioners
Built inside Makeitfuture for real company work
ChannelGate grew from the same problems we solve for clients: agents need reliable access to business systems, people need control over sensitive actions and useful methods should be reusable across the company. We build the agents and automations around it, then use ChannelGate to make that work available to the team.
- Make.com AI Partner of the Year 2025
- ISO 27001 and ISO 9001 certified
- 700+ clients and 15,000+ automations
- More than 7 years of automation delivery
Implementation
Launch with a working capability from day one
The focused ChannelGate implementation is designed to put the operating foundation in place, build the first agents and skills with your team, and prepare internal champions to continue the work.
- 30 hours over about four weeks
- Discovery, installation and standard skills
- Guided working sessions and team training
- Starting from €5,000, excluding licences, subscriptions and usage
Questions about ChannelGate
-
What is ChannelGate?
ChannelGate is a company-governed AI operating system that lets teams work with AI agents inside collaboration channels while the company controls identity, permissions, skills, memory and approvals. The product sits between the team request, the agent and the connected company systems.
-
Is ChannelGate an AI agent?
No. An AI agent performs a job, while ChannelGate provides the shared operating layer around agents. It controls how people reach them, which tools and credentials they can use, when a person must approve an action and how the work is recorded.
-
How does ChannelGate relate to Makeitfuture?
ChannelGate is a Makeitfuture product. We built it for our own AI agent work and use it as the shared AI operating layer inside the company. Makeitfuture also provides the implementation and team training described on the ChannelGate website.
-
Where can I request a licence or discuss implementation?
Product registration, implementation enquiries and enterprise discussions are handled on channelgate.dev. This Makeitfuture page explains why we built the product and how it fits with our AI services.
-
Can ChannelGate connect to our existing systems?
ChannelGate is designed to give agents approved access to company tools and connections. The right setup depends on your systems, identity model and security requirements, which are covered during implementation planning.
See ChannelGate
Ready to give your AI agents a company operating layer?
Explore the product, review the deployment and governance options, or discuss a guided implementation on the ChannelGate website.