A chatbot on your website. A branded AI workspace for your team. A local assistant on your own hardware. Tell us what it should help people do, and we’ll help choose the right setup.
Model selection · Scoped implementation · Clear running costs
These are project starting points. We define the features, access and acceptance checks with you before building.
Your own web interface for approved AI tasks, with your branding, user access and a compatible model chosen around quality, budget and the work.
For example: drafting, summarizing or working through an internal task.
Plan a team workspace ↗A chatbot that retrieves information from approved guides, procedures or product documents, with source references and access rules in the agreed scope.
For example: helping staff find a procedure and check its original source.
Discuss a knowledge assistant ↗Answer common questions, collect the right information and direct inquiries to your team. Agree when the bot should ask for help or hand off to a person.
For example: understanding a request before someone follows up.
Discuss a website chatbot ↗The right choice depends on your information, hardware, budget and the tasks the assistant needs to perform.
A branded application connects to an agreed model provider through its API.
A compatible model runs on your own workstation, server or agreed private infrastructure.
A local installation designed for agreed tasks without an external model API.
Private does not automatically mean offline. We document where data goes and which systems are involved. Retention, residency and access depend on the final configuration and the services you approve.
Online model usage can be pay as you go. Building and operating your application are scoped separately.
Bring a model preference, or let us compare compatible options with sample tasks. We agree the accounts, billing responsibilities and expected usage before launch.
Request a scoped AI quote ↗Design, configuration, integrations and agreed testing. Quoted in CAD for your scope, with applicable taxes identified.
For online APIs, the chosen provider’s usage charges and billing currency apply. We explain the arrangement and configure usage controls where supported.
Hosting, storage, domains and third-party services, or local hardware and running costs. These appear separately in the proposal.
Updates, document changes, monitoring or additional features are agreed separately. Your quote defines what is included.
No flat model rate or unlimited usage is promised. A planning budget is not a guaranteed spending cap.
If your workplace restricts AI tools, start with the business owner or IT team. We can help scope an approved deployment with the controls they need.
We agree which information may be used, who can access it, where it is processed and when a human needs to review the output.
We use AI-assisted development with human review and testing. We remain responsible for the work we deliver.
Our locally run computer vision application for tracking and counting in industrial, retail and logistics settings.
Configure objects and virtual tripwires on a camera feed, then test the setup in your environment. Discuss counting-data exports, dashboards, inventory connections and threshold alerts.
Camera position, lighting, hardware, data retention and integrations all form part of the assessment. We agree how performance will be evaluated before a rollout.
Ask about Spectre Edge ↗
Describe the task, who will use it and any model or deployment preference. We’ll clarify the fit and outline a scoped next step.
You do not need to choose the technology first. “Our staff need to find the right procedure” is a good place to start.
Keep passwords, API keys, customer records and confidential documents out of this form. We can agree a suitable way to share project material later.
Email Michael ↗Yes. We can scope a branded web workspace or chatbot with the access rules, supported tasks and integrations your team needs. Your quote identifies domain, hosting, user access, maintenance and model-provider responsibilities.
Tell us your preferred model or what you need it to do. We review availability, licensing, cost, quality and deployment requirements, then test representative tasks. A hosted model may not be available to run locally or offline, and changing models can change cost and behaviour.
For an online model connected through a provider API, model usage is billed according to the chosen provider and arrangement. Setup, application hosting, storage and ongoing support are scoped separately. We agree who owns each account, who pays each bill and how usage is measured before launch. Alerts or spending controls are configured where supported; a budget target alone is not a guaranteed cap.
No. A private workspace can restrict access while still sending requests to an external model provider. An offline installation runs a compatible local model without depending on an external model API for the agreed tasks. We review and test network behaviour, updates, logs, backups and integrations as part of the design.
Yes, by helping you assess an organization-approved deployment with your decision-maker or IT team. We agree the intended use, approved information, access and network requirements. A custom domain does not replace workplace approval or permission to process company data.
No. We test with representative questions, define what the assistant should and should not answer, and agree when it should hand off to a person. Responses may need human review. Connecting company documents does not eliminate mistakes or replace the permissions needed to access them.
Yes. A feasibility review, staff guidance or a small pilot can be scoped before a wider rollout. We start with the task and a measurable result. Sometimes a search tool, a conventional automation or a clearer process is a better fit.