What we build
- Agent architectures. Systems where specialized AI agents each handle one job and plug into a shared platform.
- Assistants grounded in your content. Answers drawn from your documentation, with guardrails that keep internal material out of customer-facing replies.
- Workflow automation. AI steps inside existing processes, such as classifying inbound files, drafting mappings or preparing records for review.
- The integration around it. APIs, data pipelines, authentication and monitoring, because an AI feature is only as good as the system it lives in.
Work we've done
For a US caregiving health platform, a Salt Square team builds the product end to end: EHR integration over FHIR, Apple Health and Google health data sync, AI health summaries and a health-assistant chatbot, plus the mobile and web apps around them.
For a healthcare AI startup, Salt Square is the core engineering partner on an agentic platform in pilot with its first customers, including releases that preserve analysts' work across data reloads.
For a hospitality technology company, we are building an AI-powered venue music platform on Google Cloud: prompt-driven playlist generation, multi-zone playback and role-based venue dashboards.
For a US digital health company, our engineers worked on an extensible agent architecture for provider scheduling, gaps in care and proactive patient outreach, on top of a platform we had supported for years.
For an enterprise software vendor, we prototyped a customer-facing support assistant on the company's internal AI platform, including the guardrails that keep internal content out of customer answers.
How we work
We build AI features the way we build any production software: small releases, tests, monitoring, and a clear owner. Where the domain is regulated, compliance and data handling are designed in from the first sprint.
Frequently asked questions
Can you work with the AI platform we already have?
Yes. We usually build on what you have rather than introduce a new stack.