AI Testing & Advisory Services

Independent evaluation for AI assistants and agents in high-stakes settings.

Request independent testing

AI Red Teaming & Adversarial Testing

  • Build realistic edge cases and multi-turn scenarios.
  • Probe for jailbreaking, prompt injection, and safeguard bypasses.
  • Test how AI handles conflicting instructions and repeated pressure.

AI Safety & Model Behavior Evaluation

  • Test for sycophancy, including excessive agreement with unsafe assumptions.
  • Look for hallucinations, unsafe advice, and inappropriate reassurance.
  • Check whether the model communicates uncertainty and escalates when needed.

AI Agent Reliability & Tool Use

  • Test tool-use boundaries, human approval points, and handoffs.
  • Check recovery when tools fail, information is missing, or the user changes direction.
  • Evaluate whether agents complete multi-step tasks within scope.

Clinical, Behavioral Health & Financial AI Safety

  • Evaluate sensitive conversations, medication guidance, and escalation.
  • Test for risky financial guidance, missed scam signals, and reinforcement of harmful choices.
  • Apply context-specific harm and severity criteria.

Evaluation Design, Rubrics & Retesting

  • Develop scenario libraries, severity rubrics, and guidance for human reviewers.
  • Document examples, severity rationale, and recommendations for what to fix first.
  • Retest selected findings after changes, where agreed.

Strategic Advising & Mentorship

  • Work through AI governance, product, or adoption decisions.
  • Get an experienced sounding board for a pilot or rollout.
  • Set up focused or ongoing mentorship for founders and leadership teams.

Pricing & Engagements

Starting prices for independent AI testing in USD.

AI Behavior & Safety Snapshot

From $750 USD

A first look at one narrow interaction or workflow, with a short findings summary and prioritized recommendations.

Focused AI Red Teaming

From $1,750 USD

Targeted adversarial testing of one assistant or workflow, with documented examples and severity-rated findings.

Independent Healthcare AI Evaluation

From $4,500 USD

A deeper evaluation of one patient-facing or clinician-support workflow, with a findings report and stakeholder discussion.

Need a broader assessment or ongoing support? We scope complex agent evaluations and ongoing testing around your goals. Final pricing depends on the workflows, system access, and depth of testing. Strategic advising and mentorship are scoped separately.

Tell us what you need to test

We can start with one assistant, agent, or workflow and agree on the behaviors that matter most.

Request a scoped quote