Skip to content

Arca Vision · machine representation

Human version

HealthTech engineering

Specialized engineering and technical architecture for organizations handling PHI. Clinical interoperability, HIPAA and SOC 2 architecture, local medical AI, and claims automation.

HEALTHTECH

  • Clinical Interoperability: HL7, FHIR, and custom EHR integrations. Interface engines between clinical systems; Proprietary EHR APIs alongside standard FHIR and HL7; A written data map before any sync ships
  • HIPAA & SOC2 Architecture: Cloud compliance, security assessments, and BAA management. Named PHI custody instead of a default cloud vault; BAAs when PHI is in scope; Security assessments and control mapping for the engagement
  • Medical AI & Local LLMs: Secure, local-first medical AI pipelines without exposing PHI. On-device or perimeter inference for clinical workloads; No default multi-tenant model API for regulated inputs; Evaluation scoped in writing
  • Claims & EDI Automation: Engineering data pipelines for healthcare claims and revenue cycles. 837 and 835 claim and remittance flows; Clearinghouse and payer interfaces; Reconciliation pipelines with a written scope

Technical capabilities

  • Edge AI & local SLMs: On-device and perimeter-bound inference when the data cannot leave a named trust boundary. We design the pipeline, evaluate the model, and implement what we can demonstrate. Local inference where a vendor API is the wrong custody model; Quantization and grammar constraints stated as engineering choices, not guarantees; Architecture, evaluation, and implementation scoped per engagement
  • Cryptographic isolation: Zero-trust infrastructure and air-gapped custody for secrets, keys, and regulated inputs. Claims stay inside a written trust boundary. Local-first secrets and key-custody architecture; Named trust boundaries instead of default cloud vaults; Systems work that can be shown, not a capability brochure
  • Custom Rust systems: Native desktop, CLI, and systems engineering in Rust when the problem needs it. Scope is agreed in writing. Product and systems engineering; Native desktop and command-line tools; Desktop, web, and systems work with a written scope

Sector deliverables

  • Commercial enterprise: We design and ship software for organizations that need it built. Government is one customer segment. Neither track is the whole company. Scoped product and platform engineering; AI systems when the problem calls for them; Written SOW, not an open-ended bench
  • Healthcare & life sciences: HIPAA-ready architecture for PHI that must remain in a named custody boundary. We do not claim a certified product; we design systems that keep clinical data off the default cloud path. Local clinical structuring and ambient workflows we will actually take; BAAs when PHI is in scope; No default cloud vault for regulated inputs

Government is a customer segment, not the brand

We respond to government RFPs, including through SAM.gov. Identifiers, NAICS, and the printable capability statement live on a dedicated procurement page.

SAM.gov UEI, D-U-N-S, NAICS, socio-economic designations, and the one-page capability statement are on /government. Award-specific past performance is provided in proposals.

Links