AI / ML Platforms
AI and machine learning platforms, model access, and MLOps tooling. We also build custom AI on these APIs — see AI Solutions.
How ai / ml platforms is bought
What agencies usually ask for
AI platform purchases follow a pilot that worked: a document or records assistant, a public-facing help experience, or an analytics model that now needs a production home. They also follow governance, when an agency decides which platforms staff may use with agency data and wants an enterprise agreement that states how that data is handled.
How this category is licensed
Managed model platforms such as Amazon Bedrock and Vertex AI are consumption: priced per unit of text processed, by model, with provisioned throughput available for predictable workloads. NVIDIA AI Enterprise is a per-GPU subscription for agencies running models on their own servers. Data handling terms, the region the service runs in, and whether your data is used for training are contract points to confirm in writing before the quote is accepted.
What to have ready for a quote
The use cases and their expected volume, the sensitivity of the data involved and the compliance regime that applies, which cloud you already run, and whether you need the application built as well as the platform licensed. Consumption is quoted as an estimate with the assumptions written down; the platform, any GPU hardware, and build work are separate lines.
Typical combinations
Amazon Bedrock inside an existing GovCloud landing zone, or Vertex AI alongside Google Workspace, keeps the platform inside a boundary the agency already governs. NVIDIA AI Enterprise pairs with HPE or Dell servers for on-premises inference. Custom applications on any of these APIs are built by our AI Solutions team and quoted as a project.
What moves the price most
Which model handles each task, since the largest models cost several times more per unit than smaller ones that are often sufficient; provisioned throughput versus on-demand; the volume of text processed; the number of GPUs for on-premises inference; and the region. Because consumption estimates depend on usage assumptions, we state those assumptions on the quote so they can be challenged.
AI / ML Platforms
across every line
A selection of the ai / ml platforms lines we source most often. We quote well beyond what is listed here — if you do not see a product, ask and we will price it.
Vertex AI Platform
Managed platform to build, tune, and deploy generative-AI models.
AI Enterprise
Optimized AI/ML frameworks and accelerated-compute software stack.
Amazon Bedrock
Build and scale generative-AI apps with foundation models, securely.
Need this quoted?
Add the lines you're sourcing and send them over. We respond within one business day with a consolidated quote and the right contract vehicle.