Skip to main content
Software SushiRequest a fit call

HEALTHCARE AI ENGINEERING

AI systems for teamsthat handle PHI.

Healthcare AI development for health-tech teams working under HIPAA: LLM features, clinical-workflow AI and document automation, with PHI handling designed in from day one. Start with a fixed-scope pilot or a dedicated team.

Let’s talk healthcare AI.

Share your goals. We’ll arrange a fit call.

Please use your company email address.
More details (optional)

Please don’t include PHI, patient information, credentials or other sensitive data in this form.

Fixed price for pilots. Monthly retainer for teams.

By submitting, you agree to our privacy policy.

Client ratings

Clutch5.0/53 reviews
Upwork5.0/513 ratings

We partner with

01 / THE SCOPE

Start with one useful workflow.
Designed for careful review.

Building a HIPAA-compliant LLM feature? Compliance is your program. Our job is to build systems that fit it and make your security review straightforward. Bring us the healthcare workflow you want to improve. We agree on the data, success criteria and boundaries before work begins.

01

One focused workflow

LLM features, document automation or clinician-led review, scoped around a clear need.

02

A bounded data sample

Synthetic or de-identified data for an initial pilot, with agreed access boundaries.

03

A working prototype

A tangible starting point in a non-production environment.

04

An evaluation set

Representative cases to assess usefulness and limitations.

05

Controls designed in

Plan data minimization, access, audit records and human review.

06

A path forward

A results report and a recommendation for production work or a dedicated team.

BOUNDARIES, UP FRONT

Clear data. Clear responsibilities.

  • No PHI in this form
  • BAA when work involves PHI
  • Pilot uses synthetic or de-identified data
  • No autonomous clinical decisions

02 / EXPERIENCE BEHIND THE WORK

We’ve built beyond the demo.

Healthcare work calls for thoughtful engineering, controlled data access and a clear place for professional judgment.

CLINICIAN-LED REVIEW

Risk estimates.
Professional decisions.

Prediction models, an API and assessment interfaces that keep clinician decisions separate from model estimates.

ANONYMIZED PROJECT EXAMPLE

HEALTH DATA

Complex data.
Careful engineering.

Engineering for a health AI platform working with PHI and CJIS data.

ANONYMIZED PROJECT EXAMPLE

DOCUMENT INTELLIGENCE

Less manual work.
More review context.

Our document engineering experience informs retrieval, classification and review workflows.

ENGINEERING EXPERIENCE

Previous project experience. These examples do not promise outcomes or imply compliance certification. Read the clinician-led case study ↗

TEAM CREDENTIALS

Certified in the tools we build with.

AWS Certified AI Practitioner
Claude Certified Developer – Foundations
AWS Certified Solutions Architect – Professional

03 / HOW IT WORKS

From a useful question
to a reviewed next step.

Request a fit call
  1. 01

    Fit call

    Map the workflow, data boundaries and what you want to learn.

  2. 02

    Agreed proposal

    Choose a fixed-scope pilot or dedicated team, with deliverables and pricing.

  3. 03

    Build + review

    Work in the agreed environment, with evaluation and human review.

  4. 04

    Next steps

    Review the evidence and scope production delivery separately.

Two ways to start. Test one use case with a fixed-scope pilot, or bring on a managed team for ongoing roadmap delivery. Explore dedicated teams ↗

BEFORE WE TALK

Good questions.
Straight answers.

What your product, security and legal teams need to know.

01Do you work with teams under HIPAA?

Yes. We design for your HIPAA program: data minimization, redaction before any data reaches an LLM, role-based access and audit logging, deployed in your environment where possible. We sign a BAA when our work involves PHI. There’s no official “HIPAA certification” for vendors, so instead we support your security review and complete your questionnaire.

02Will you sign a BAA?

Yes. We sign a Business Associate Agreement for engagements where we create, receive, store or access PHI on your behalf. We agree on the agreement and security requirements before work begins.

03Are you SOC 2 certified?

Not currently. We support your security questionnaire and review the project's controls with you. Work in your environment and the policies required for the engagement are agreed before kickoff.

04Where does our data go when you use an LLM?

We agree on data flows, providers and configurations with you before implementation. Options can include minimization, redaction, approved hosted services or self-hosted models. Retention and BAA eligibility must be checked for the selected service and configuration.

05Who owns the code and models?

Code ownership and handoff are defined in the proposal before you commit. We identify custom deliverables, pre-existing components, and third-party or open-source licenses so you know exactly what you receive.

06Do you work with US-based staff?

Tell us your location, access and time-zone requirements on the fit call. We confirm the staffing arrangement in your proposal.

07What does the pilot cost?

A fixed price, quoted after the fit call. We agree on one use case, a bounded synthetic or de-identified data sample, deliverables and evaluation criteria.

CLIENT PERSPECTIVES

Good work.
Real relationships.

Upwork · 5.0/5 · 13 ratings
Includes ratings without comments.
Selected written feedback from Clutch and Upwork below.

Feedback from different Software Sushi projects. Original sources and project types are shown for context.

A CLEAR NEXT STEP.

Have a healthcare workflow in mind?

Tell us what you want to build. We’ll discuss the fit and define the next step.

Request a fit call

Platform partnerships

Book a call

Schedule a free consultation to discuss your project with an engineer.