Enterprise-grade AI · Regulated industry · SMB scale

Enterprise-grade AI quality for the SMB and mid-market sectors that normally can't afford it.

Independent AI assessment and governance for the places where a wrong answer becomes a filing.

The work is built to the standards regulated industry already runs on: FDA 21 CFR Part 11, HIPAA, CMMC, GxP, and the EU AI Act.

Book a 30-minute call See the practice
  • 25+ years in quality engineering
  • Fortune 1-100 scale QE organizations
  • Clinical, finance, and banking environments

AI Readiness Assessment

Redacted · mid-market · regulated

Sample
  • Infrastructure and access 3/4
  • Governance and risk 4/7
  • Operational reliability 4/6
  • Automation opportunity 3/4
  • Readiness and culture 2/3

24 dimensions across five categories

Musings

Notes on the OpenAI and Hugging Face Incident

A recap of what happened, what it means for teams running open models, and what to check this week.

Read
A guard that validates the wording of a promise is not a guard on the promise. An alert that outlives its own fix teaches you to scroll past alerts. Absence of evidence rendered as evidence of absence is how a monitor lies politely.
All musings and briefs

What iSL does

One practice, applied where the rules are hardest.

AI advisory is the practice. eCOA/IRT and finance are where it is applied, each with its own regulator and its own evidence bar: a claim ships only when something machine-checkable backs it.

AI Advisory

Know whether the AI is safe to act on, before you act on it.

Independent assessment and governance for businesses adopting AI in an environment where a wrong answer carries a regulatory consequence.

  • Scored across 24 dimensions in five categories: infrastructure and access, governance and risk, operational reliability, automation opportunity, and readiness.
  • Two engagement tiers: Vector for a first AI implementation, Meridian for a business with AI in place and ready to do significantly more.
  • Fortune 100 governance frameworks, run at the speed a smaller business can actually absorb.
Learn more

Applied in

eCOA/IRT

Clinical outcomes assessment that holds up under inspection.

Validation for electronic clinical outcome assessment data in regulated trial environments.

  • Designed to a regulatory superset: 21 CFR Part 11, GxP, HIPAA, GDPR, and the EU AI Act.
  • Runs local and air-gapped, so PHI stays inside the environment it belongs to.
  • A fail-closed grounding gate: an answer ships only when a valid citation backs it, with an adversarial self-test battery so it cannot quietly regress.
Learn more

Finance Advisory

Audit-ready books, and a CFO's read on what they mean.

Finance advisory for lending, underwriting, and reporting workflows that sit under regulatory scrutiny.

  • Starts with audit-ready books and clean reporting, then grows into strategic CFO access: runway, pricing, and growth decisions.
  • Built for SMB owners, commercial clients, and non-profits carrying fund and grant accounting.
  • AI-augmented delivery, with regulated attest work carried by the licensed professional who signs it.
Learn more

Regulated-industry pedigree

The standards this work was learned in.

Regulated industry does not grade on effort. These are the frameworks and environments behind the practice, stated plainly.

FDA 21 CFR Part 11

Electronic records and electronic signatures, including the validation and audit-trail discipline that comes with them.

HIPAA and PHI handling

Protected health information treated as in-scope by default, with de-identification and access control designed in rather than added after.

CMMC

Cybersecurity maturity expectations for organizations working inside a defense supply chain.

GxP and Computer Software Assurance

Risk-based assurance for life sciences systems, where the depth of testing has to be justified, not assumed.

EU AI Act

Obligations that arrive with a system's risk classification, tracked as the rules take effect.

25+ years of quality engineering

Quality engineering and release management organizations built at Fortune 1-100 scale, across clinical, life sciences, finance, banking, retail, and global technology.

Published work

An independent guide, every claim source-linked.

The AI Liability Insurance Guide, 2026 edition

A buyer's guide to a market with almost no loss history: what the policies actually cover, where the exclusions are moving, and what underwriters ask before they quote. Written and maintained by the practice, with every claim linked to its source.

Visit the guide

Also from the practice

Three more ways we work with you.

Web Presence as a Service

A digital home built right and kept that way: design, hosting, and ongoing care.

Explore

Fractional C-suite

Senior technology leadership on demand, with Fortune 1-100 engineering rigor.

Explore

Quality Engineering

Test strategy, automation, release discipline, and QE for AI-generated code.

Explore

Let's talk

The right fit starts with a conversation.

Thirty minutes, no preparation needed: you describe where AI already touches your business, and you leave with a plain read on what is safe to act on now and what needs evidence first. Every engagement is scoped to what you actually need, so the number follows the conversation rather than the other way around.

Book a 30-minute call Read the briefs first