Complete AI Training
Sign inGet my AI kit

Your job's AI kit

Get your AI kit

Tell us who you are and what you do. We show you your kit right away and email you the link: skills, prompts, AI agents, MCP servers and courses for your job.

500+ jobs ready, and we make a kit for any other job. No payment needed to look.

Share

AI agent for biomedical engineers

Device Software Change Assessment Agent

A documented impact assessment and test plan for each software change, with no requirement left untested.

ComplianceDecisionsCreatingScience and researchProduct development
See how it worksGet access
Device Software Change Assessment Agent: what goes in, what the agent does and what you get

What it does

A developer says it is a small fix, and then someone has to decide whether it needs a new regulatory filing and which tests to rerun. This agent reads the change description and the code diff summary, then traces which requirements, risk controls and tests are affected. It drafts an impact assessment covering safety classification, changes to intended use, and the regression tests needed. It checks that every affected requirement has a test and that every risk control touched is verified, and loops to propose added tests where something is missing. It lists open questions for you. You approve the decision on whether the change needs a regulatory submission. Edge case: a change in a software library is assessed for every function that uses it.

How it works

Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.

Start and resultWhat it doesA check on its own workWaits for your OKGoes back and retries
Yes, continueYes, continueApprovedNoNo 1 STARTS WHEN Software change submitted 2 USES A TOOL Read the change description and diff summary 3 USES A TOOL Trace affected requirements, risk controls and testsin the matrix 4 DOES Rate impact on safety classification and intendeduse 5 CHECKS THE RESULT Does every affected requirement have a test? If not: propose new or updated tests and add them to theplan. Back to step 3. 6 CHECKS THE RESULT Is every touched risk control verified by a test? If not: list the verification gaps and add them to theplan. Back to step 3. 7 DOES Draft the impact assessment and regression test list 8 DOES List open questions and assumptions 9 YOU APPROVE Engineer approves the regulatory decision 10 USES A TOOL Record the assessment as a draft in the qualitysystem 11 RESULT Impact assessment with test plan
Read the steps as a list
  1. Software change submitted
  2. Read the change description and diff summary
  3. Trace affected requirements, risk controls and tests in the matrix
  4. Rate impact on safety classification and intended use
  5. Does every affected requirement have a test?If not: propose new or updated tests and add them to the plan. Back to step 3.
  6. Is every touched risk control verified by a test?If not: list the verification gaps and add them to the plan. Back to step 3.
  7. Draft the impact assessment and regression test list
  8. List open questions and assumptions
  9. Engineer approves the regulatory decisionThe agent waits here for your OK.
  10. Record the assessment as a draft in the quality system
  11. Impact assessment with test plan

How it decides

It traces the change through the trace matrix. Changes touching safety-related items, intended use or algorithms are marked as higher impact and need the engineer's regulatory judgment.

  • Change touches an alarm or dosing function: high impact, full regression
  • Change alters an algorithm output: ask for a regulatory review
  • UI text only with no safety link: low impact, targeted tests
  • Library update: assess every function that depends on it

Make it yours

Every agent is a starting point. You choose these settings for your own situation.

  • Impact categories
  • Regression test sets by area
  • Classification scheme
  • Quality system templates
  • Reviewers

What keeps you in control

It always asks you first

  • The regulatory impact decision
  • The final test plan

Hard limits

  • Never decide that a change needs no regulatory action
  • Never edit the trace matrix without approval

It stops when

  • Done: assessment approved and tests listed
  • Stop: change description too vague, ask the developer

Set it up

We guide you through the set-up, step by step

Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.

10 minto set it up in your AI
5 AIsChatGPT, Claude, Copilot, Gemini, Grok
  • One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
  • The agent then walks you through connecting your own data, one source at a time
  • A downloadable copy with the flow chart, the rules and the full guide
Get access to this agent

An example run

What happensA change updated a filtering routine used by the pressure alarm. The agent traced 4 requirements and 2 risk controls. One requirement, R-118 on alarm delay, had no test, so the check failed. The agent proposed a test with a 3 second tolerance and a boundary case. It marked the change high impact. The engineer approved the plan and decided no new submission was needed.

More agents for biomedical engineers

Communication · Automation

Product Requirement Change Impact Agent

Every requirement change reaches all linked tests, docs and teams.

7 steps1 check3 jobs
Analysis · Decisions

User Test Prototype Iteration Agent

Turn test observations into ranked problems and verified changes for the next prototype.

11 steps2 checks8 jobs
Project tracking · Administration

Lab Meeting Action Follow-up Agent

Make sure every action item from a meeting has an owner, a date and a recorded outcome.

10 steps2 checks7 jobs
Risk management · Quality control

Usability Study Use Error Tracking Agent

Every observed use error, close call and difficulty is traced to a hazard and a mitigation before the report.

11 steps2 checks2 jobs