Partners & platforms
The platforms we build on.
These are the platforms we build on, deploy to, and integrate.
How to read this
All platform today.
We've grouped the technology by what it does, not by who we know. Each group describes what that category means for your build: we name the platforms in it, and we deploy whichever one fits the job. None of these listings implies a relationship or an endorsement of us; it's the technology, named honestly.
Named honestly, with any interest disclosed below and in our principles.
Foundational AI models
The models the AI work runs on.
When an LLM is the right tool (for triage, knowledge, agent assist or the AI work tied to a metric you own) this is where it comes from. We pick the model that fits the task, the budget and any data-residency requirement, and we say when an LLM isn't the answer at all. Some providers ship their own models too: Microsoft, for example, builds its own (such as Phi), hosted on the major clouds (AWS, Google and others).
- Anthropic
- OpenAI
- Mistral
- Z.ai
- Microsoft
Cloud & infrastructure
Where the build runs.
The cloud estates we host, deploy and run workloads on. We fit the build to the tooling, identity and data residency you already have, so it sits inside the platform your team already runs rather than bolting on a new one.
- AWS
- Microsoft
CX & support-platform ecosystem
The support platform and what extends it.
The support platform we set up, clean up, migrate and tune most often, plus the ecosystem apps that extend it. We configure these to fit how your team actually works, and we tell you where the native feature already does the job before adding anything.
- Zendesk
- Stylo
- agnoStack
- CloudSet
Voice & contact centre
Calls, routing and natural speech.
The platforms we build call flows, routing, IVR and contact-centre workflows on, including natural speech where it's part of the service. We design the journey around the outcome you're measured on, not around any one vendor's feature list.
- Twilio
- Amazon Connect
- ElevenLabs
Data & analytics
Data, dashboards and AI analytics.
Where we build the data layer and the analytics on top: warehouse and lakehouse on Microsoft Fabric with OneLake, Azure Synapse Analytics and Amazon Redshift; query and ETL with Amazon Athena and AWS Glue; and the BI and dashboards your team actually watches, from Power BI and Amazon QuickSight to Geckoboard. It also covers the AI-analytics stack: receipt and document AI (Amazon Textract, Azure AI Document Intelligence), intent and entity extraction (Amazon Comprehend, Azure AI Language), fraud detection (Amazon Fraud Detector), and model training and serving (Amazon SageMaker AI, Azure Machine Learning). We build it so the metric you own is something you can see and trust, not a figure we report back to you.
- Microsoft Fabric
- Power BI
- Amazon Redshift
- Amazon QuickSight
- Amazon SageMaker AI
- Amazon Textract
- Geckoboard
Language & translation
When language blocks service quality.
Real-time translation we integrate where language gets in the way of service quality, leaning toward voice. One enabler of multilingual support, not the whole story.
- DeepL
Commercial interest
Where we have a commercial interest, we say so.
You buy any platform direct from the supplier. We recommend what works for your situation, and we name any interest we hold (including any formal partner status, should we form one) so you can weigh the advice yourself.