Question 1

What does the Private AI service include?

Accepted Answer

Design, deployment, and ongoing management of a bespoke self-hosted AI system: model selection, hosting architecture, a controlled chat interface, sign-in and access groups, logging, and documented operating boundaries. After go-live, Teclara keeps managing the system, including model updates, monitoring, and tuning. The result is drafting, summarization, and research support over the files you choose to include, running on infrastructure your firm controls.

Question 2

What infrastructure does this run on?

Accepted Answer

Two options. Dedicated on-premises hardware, which you can supply or we can procure for your office, or a private cloud environment with rented GPUs running your private models. We confirm sizing and the right option during the scope review, and will tell you honestly if a setup will not carry the workload. Our pricing covers the design, build, and ongoing management.

Question 3

How does our client data stay private?

Accepted Answer

The model runs on infrastructure you control, so prompts and files never reach public AI services and no client data trains public models. Access follows the same permissions that already protect your files, and every prompt, query, and output is logged and reviewable.

Question 4

Which models and platforms do you work with?

Accepted Answer

Open-weight models selected for the work you need, running on dedicated on-premises hardware or on rented GPUs in a private cloud. The deployment sits alongside Microsoft 365 or Google Workspace and uses your existing identity for sign-in. If your requirements point somewhere we do not work, we will say so during discovery.

Question 5

How does an engagement start?

Accepted Answer

With a scope review: data sensitivity, target users, infrastructure constraints, and what the system needs to prove. From there we design the architecture, build the deployment, and validate it with your team. The system then runs as a managed service, with Teclara handling model updates, monitoring, tuning, and support, backed by owner documentation and usage guidance.

Question 6

Can Teclara deliver Private AI in the United States?

Accepted Answer

Yes. Design, deployment, and ongoing management are delivered remotely from Canada. Private cloud deployments on rented GPUs are fully remote. On-premises hardware for a U.S. office is scoped per engagement, including procurement and local logistics.

Question 7

Can U.S. clients be invoiced in U.S. dollars?

Accepted Answer

Yes. Teclara can invoice U.S. clients in USD and receive payment through U.S. banking.

Question 8

Where does the data live?

Accepted Answer

On infrastructure your firm controls, either dedicated hardware or a private cloud tenancy. No client data trains public models, and access follows the permissions your firm already uses.

AI Built for Businesses That Can't Afford a Data Leak