Platforms we have built

11 production AI systems shipped across healthcare, wellness, wealth management, manufacturing, fitness, and consumer apps. Every build was architected and shipped by Colter Mahlum personally, from initial data audit through production deployment and MLOps handoff.

Last updated: September 8, 2026

How to use this portfolio

These case studies are project snapshots, not interchangeable packages. Read each one as an account of one specific business context. Look at the operating problem, the systems involved, and the work needed to make the result useful after launch.

The industries and use cases vary. The decision process behind them is the same every time:

• Define the outcome you want.
• Understand the data available to support it.
• Map the integrations required.
• Agree on who will use the result day to day.

If a project resembles your situation, open the full case study for implementation detail. If it does not, use it to frame a sharper discovery question about data access, workflow ownership, change management, or the evidence you need before a wider rollout. That fit matters more than matching a project to a generic technology label.

What each case study helps you evaluate

• Business context. The operational or customer-facing problem the platform was built to address.
• Delivery scope. The data, integrations, model work, and workflow design involved in the build.
• Decision relevance. The outcome and implementation details a team can use when evaluating a similar initiative.

What production means

These are not prototypes or demos. Each platform listed below runs on live data, serves real users, and is built to the engineering standard its industry requires:

• Healthcare and wellness platforms are architected to HIPAA-aligned data handling requirements.
• Financial services platforms include compliance-grade audit trails and model documentation.
• Manufacturing and operations platforms run on production sensor and transactional data, not benchmark datasets.

How we select engagements

Mahlum Innovations takes a small number of custom platform engagements each year so every build gets direct founder attention. An engagement gets selected when three things are true: the business problem is clearly defined, the data exists to support the solution, and the client has executive buy-in to act on the output. We say so at the discovery call rather than start a project likely to stall. The free AI readiness assessment benchmarks your data maturity and strategic alignment before we talk.

Engineering standards and delivery model

Every platform in this portfolio is built to the engineering standard its industry requires:

• Healthcare and wellness platforms use HIPAA-aligned data handling, end-to-end encryption, role-based access control, and audit trails.
• Financial services and wealth management platforms include compliance-grade model documentation and audit-ready data lineage.
• Manufacturing and operations platforms run on live sensor and transactional data — not benchmark datasets — and are validated in production conditions before handoff.

Mahlum Innovations is a one-engineer firm by design. Colter Mahlum architects, builds, and delivers every platform personally. Nothing gets lost in handoffs, no junior engineer makes scope decisions, and the founder knows the problem domain as well as the technical stack. A limited number of custom engagements run in parallel — that limit is intentional.

AxiomAI — AI employees for recurring workflows

Not every business problem needs a custom build. For recurring workflows, the AxiomAI catalog offers purpose-built AI employees across every business function:

• Aria SDR for outbound sales
• Care CX for patient and customer communication
• Ledger for financial reporting
• AxiomCMO for AI visibility and content publishing
• HIPAA Guard for compliance monitoring
• Sage Support for internal helpdesk automation

These are subscription-based and deployable without a custom build.

Start a conversation

If you are evaluating a custom AI platform build, the free AI readiness assessment benchmarks your data maturity, technical readiness, and strategic alignment before we talk. The first conversation is free and takes 45 minutes. Book a strategy call →