
How we work
A clear engagement model, from first conversation to ongoing support.
We follow a client-owned infrastructure model. Your organisation keeps ownership of its cloud environment and billing; we provide the engineering.
Process
Six steps. No surprises.
Step 1
Discovery
A business contacts us and explains its requirements. We arrange a consultation to understand:
- Business processes
- Existing applications
- Operational challenges
- Integration requirements
- Security and access requirements
- Desired outcomes
Step 2
Assessment and proposal
We review the requirements and determine whether an AI or automation solution is appropriate. We prepare:
- Proposed architecture
- Project scope
- Deliverables
- Timeline
- Engineering fee
- Estimated third-party infrastructure requirements
- Support options
Step 3
Client infrastructure
The client owns and controls its cloud environment. For Google Cloud projects:
- The client owns the Google Cloud organisation or account
- The client owns the project
- The client owns the billing account
- The client attaches its own corporate payment method
- The client authorises appropriate engineering access
- The client's IT administrator can participate in the process
We use least-privilege permissions and approved access procedures. We will never ask you to send passwords or service account private keys over WhatsApp or ordinary email.
Step 4
Engineering
We implement the agreed solution in the client's authorised environment. Possible technologies include:
- Google ADK
- Vertex AI
- Gemini
- Python
- APIs
- Microsoft 365
- Google Workspace
- Suitable automation platforms
Technology choices depend on the project.
Step 5
Testing and deployment
We validate:
- Functionality
- Security controls
- Permissions
- Integrations
- Error handling
- Performance
- Agreed acceptance criteria
We deploy the solution after the agreed testing process.
Step 6
Support and maintenance
We provide support according to a separately agreed arrangement, which defines:
- Support availability
- Response times
- Included maintenance
- Bug-fix responsibilities
- Monitoring responsibilities
- Additional development charges
- Change request procedures
Support is always scoped — we do not offer unlimited support.
Commercial model
Transparent engineering. Independent infrastructure.
Three separate financial components, so you always know who you are paying and for what.
Client-owned environment
Owned and billed by your organisation
Google Cloud / Microsoft / Other platforms
Haki Systems engineering
Engineering fee, quoted by scope
Agents / Integrations / Automation / Testing
Business operations
Your teams, your control
Deployed systems / Workflow improvements / Ongoing support
01
Engineering and implementation
Paid to Haki Systems
The client pays Haki Systems for agreed engineering, development, integration and deployment work. Fees are quoted according to scope and complexity.
02
Third-party infrastructure
Paid directly to providers
The client pays Google Cloud, Microsoft, AI providers and other third-party vendors directly. These costs are not included in our engineering invoice unless explicitly agreed otherwise. Infrastructure and API usage is not free.
03
Ongoing support
Quoted separately
Maintenance and support are quoted separately. The service agreement defines what is included and what counts as additional development.
FAQ
Common questions.
Let's engineer what's next.
Have a business challenge that could benefit from intelligent automation or custom AI systems? Tell us what you are trying to achieve. Let's explore the possibilities.
Start a Conversation