The Omnipod GO Insulin Delivery Device is intended for the subcutaneous infusion of insulin at a preset basal rate in one 24-hour time period for 3 days (72 hours) in adults with type 2 diabetes.
Device Story
Omnipod GO is a single-use, sterile, disposable, tubeless insulin pump worn on-body via adhesive pad. Device delivers insulin at a fixed basal rate over 72 hours; no bolus capability. User fills reservoir using fill accessory. Electromechanical step-drive mechanism, activated by microprocessor, turns leadscrew to press on syringe-style reservoir, delivering insulin through integrated 27-gauge stainless steel needle and flexible cannula. Device provides audible and visual (LED) status notifications, alerts, and alarms. Occlusion algorithm monitors for increased resistance. Used by patients in home environment. Healthcare provider prescribes device; patient self-operates. Output (insulin delivery) manages blood glucose levels in type 2 diabetes patients.
Clinical Evidence
No clinical data. Evidence consists of bench testing, including basal delivery accuracy (12 pumps tested at low, medium, high rates), occlusion detection (21 pods tested), biocompatibility (leveraged from K211575, plus additional off-gas/cytotoxicity testing), human factors validation, and electrical/wireless safety testing.
Technological Characteristics
Wearable, battery-powered, single-use infusion pump. Materials identical to Omnipod DASH (K211575). Sensing/actuation: automated needle/cannula deployment and fixed-rate basal delivery. Dimensions/form factor: on-body pod with adhesive. Connectivity: none (standalone). Sterilization: Ethylene Oxide (EO). Software: embedded firmware for alarm management and delivery control.
Indications for Use
Indicated for subcutaneous insulin infusion at preset basal rates in adults with type 2 diabetes. Contraindicated for individuals unable to monitor glucose, maintain contact with healthcare providers, follow device instructions, or recognize audible/visual alerts due to hearing/vision impairment.
Regulatory Classification
Identification
An infusion pump is a device used in a health care facility to pump fluids into a patient in a controlled manner. The device may use a piston pump, a roller pump, or a peristaltic pump and may be powered electrically or mechanically. The device may also operate using a constant force to propel the fluid through a narrow tube which determines the flow rate. The device may include means to detect a fault condition, such as air in, or blockage of, the infusion line and to activate an alarm.
{0}
FDA
U.S. FOOD & DRUG
ADMINISTRATION
# 510(k) SUBSTANTIAL EQUIVALENCE DETERMINATION DECISION SUMMARY
## I Background Information:
A 510(k) Number
K223372
B Applicant
Insulet Corporation
C Proprietary and Established Names
Omnipod GO Insulin Delivery Device
D Regulatory Information
| Product Code(s) | Classification | Regulation Section | Panel |
| --- | --- | --- | --- |
| LZG | Class II | 21 CFR 880.5725 - Infusion Pump | CH - Clinical Chemistry |
E Purpose for Submission:
New device
## II Intended Use/Indications for Use:
A Intended Use(s):
See Indications for Use below.
B Indication(s) for Use:
The Omnipod GO Insulin Delivery Device is intended for the subcutaneous infusion of insulin at a preset basal rate in one 24-hour time period for 3 days (72 hours) in adults with type 2 diabetes.
C Special Conditions for Use Statement(s):
This device is for prescription use only.
Insulin pump therapy is NOT recommended for people who:
- are unable to monitor glucose as recommended by their healthcare provider.
Food and Drug Administration
10903 New Hampshire Avenue
Silver Spring, MD 20993-0002
www.fda.gov
{1}
- are unable to maintain contact with their healthcare provider.
- are unable to use the Omnipod GO Pod according to instructions.
- do NOT have adequate hearing and/or vision to allow recognition of Pod lights and sounds that signify alerts and alarms.
The Pod must be removed before Magnetic Resonance Imaging (MRI), Computed Tomography (CT) scan, and diathermy treatment. Exposure to MRI, CT, or diathermy treatment can damage the Pod.
The device is designed to use rapid-acting U-100 insulin. The Omnipod GO Pod is compatible with the following U-100 insulins: NovoLog, Fiasp, Humalog, Admelog, and Lyumjev.
## III Device Description
The subject device is intended for the subcutaneous infusion of insulin at a preset basal rate (i.e. for a fixed volume in one 24-hour time period) for 3 days (72 hours) in adults with type 2 diabetes. The subject device will have seven different models: 10, 15, 20, 25, 30, 35, and 40 units per day, with each model delivering insulin at a fixed rate over the 72-hour use life of the device. The subject device does not offer the ability to deliver a bolus.
The subject device will be entirely self-contained in an on-body pod that is single-use, sterile, and disposable. It is designed to be attached to the body via an adhesive pad that is heat-staked to the bottom of the pod. The adhesive backing keeps the pod securely in place for up to 3 days (72 hours). The sterilized package of the subject device includes the Omnipod GO pod, a fill syringe, and a fill needle.
The subject device will operate without a controller device such as a Personal Diabetes Manager (PDM), and it will automatically activate when the user fills the reservoir with insulin using the included fill needle and fill syringe. The user has three minutes to apply the pod before the needle and cannula automatically deploy. The device will provide users with visual and audible notifications related to pod set up/insulin fill, activation of basal rate, expiration, and hazardous or error scenarios resulting in deactivation.
## IV Substantial Equivalence Information:
A Predicate Device Name(s):
Valeritas V-Go Insulin Delivery Device
B Predicate 510(k) Number(s):
K103825
K223372 - Page 2 of 9
{2}
K223372 - Page 3 of 9
# C Comparison with Predicate(s):
| Device & Predicate Device(s): | K223372 | K103825 |
| --- | --- | --- |
| Device Trade Name | Omnipod GO Insulin Delivery Device | Valeritas V-Go Insulin Delivery Device |
| General Device Characteristic Similarities | | |
| Intended Use/Indications For Use | Intended for the continuous subcutaneous infusion of insulin at a preset basal rate in adults requiring insulin. | Same |
| General Device Characteristic Differences | | |
| Specific Drug/Biologic Use | U-100 Insulin.
System has been tested with NovoLog, Fiasp, Humalog, Admelog, and Lyumjev. | U-100 Insulin.
System has been tested with NovoLog and Humalog. |
| Insulin Delivery Modes | Basal only | Both basal and bolus |
| Preset Basal Rates | 10, 15, 20, 25, 30, 35, and 40 U/day | 20, 30, and 40 U/day |
| Duration of Use | Up to 72 hours | Up to 24 hours |
# V Standards/Guidance Documents Referenced:
- ISO 10993-1 2018/Cor.1: 2010, Biological Evaluation of Medical Devices - Part 1: Evaluation and Testing within a Risk Management Process
- ISO 10993-3: 2014, Biological Evaluation of Medical Devices - Part 3: Tests for Genotoxicity, Carcinogenicity and Reproductive Toxicity
- ISO 10993-5: 2009, Biological Evaluation of Medical Devices - Part 5: Tests for in Vitro Cytotoxicity
- ISO 10993-6: 2016, Biological Evaluation of Medical Devices - Part 6: Test for Local Effects After Implantation
- ISO 10993-7 2008/Cor 1: 2009, Biological Evaluation of Medical Devices - Part 7: Ethylene Oxide Sterilization Residuals
- ISO 10993-10: 2010, Biological Evaluation of Medical Devices - Part 10: Tests for Irritation and Skin Sensitization
{3}
- ISO 10993-11: 2017, Biological Evaluation of Medical Devices - Part 11: Tests for Systemic Toxicity
- ISO 10993-12: 2021, Biological Evaluation of Medical Devices - Part 12: Sample preparation and reference materials
- ISO 10993-17: 2002, Biological Evaluation of Medical Devices - Part 17: Methods for the establishment of allowable limits for leachable substances
- ISO 10993-18: 2020, Biological Evaluation of Medical Devices - Part 18: Chemical characterization of materials
- ISO 11135 2014+A1: 2018, Sterilization of health care products - Ethylene oxide - Requirements for development, validation and routine control of a sterilization process for medical devices
- ISO 11737-1: 2018, Sterilization of health care products - microbiological methods - part 1: Determination of a population of microorganisms on product
- ISO 11607-1: 2019, Packaging for terminally sterilized medical devices - Part 1: Requirements for materials, sterile barrier systems and packaging systems
- ISO 11607-2: 2019, Packaging for Terminally Sterilized Medical Devices - Part 2: Validation requirements for forming, sealing and assembly processes
- IEC 60601-1-2 Ed. 4.0: 2014, Medical Electrical Equipment- Part 1-2: Collateral Standard: Electromagnetic Disturbances - Requirements and Tests
- ISTA 3A: 2018, Packaged-Products for Parcel Delivery System Shipment 70 kg (150 lb) or Less
- ASTM F1980-16: 2016, Standard Guide for Accelerated Aging of Sterile Barrier Systems for Medical Devices
- ASTM D4169-16: 2016, Standard Practice for Performance Testing of Shipping Containers and Systems
- ASTM F88/F88M- 15: 2015, Standard Test method for Peel Strength of Flexible Barrier Materials
- ASTM F2096-11: 2019, Standard Test Method for Detecting Gross Leaks in Packaging by Internal Pressure (Bubble Test)
- ASTM F1929-15: 2015, Standard Test Method for Detecting Seal Leaks in Medical Packaging by Dye Penetration
- F1886/F1886M-16: 2016, Standard Test Method for Determining Integrity of Seals for Flexible Packaging by Visual Inspection
- ASTM D903-98: 2017, Standard Test method for Peel of Stripping Strength of Adhesive Bonds
- IEC 60601-1-6 Ed. 3.1: 2013, Medical Electrical Equipment - Part 1-6: Collateral standard: Usability
- IEC 62366-1 Ed. 1: 2015, Medical Devices – Part 1: Application of usability engineering to medical devices
K223372 - Page 4 of 9
{4}
- ISO 15223-1 Ed. 3: 2016, Medical devices - Symbols to be used with medical device labels, labelling and information to be supplied - Part 1: General requirements
- ISO 14971 Ed. 2: 2019, Medical Devices - Application of Risk Management to Medical Devices
- IEC 60601-1 Ed. 3.1: 2012 Medical Electrical Equipment - Part 1: General Requirements for Basic Safety and Essential Performance
- ANSI/AAMI ES60601-1: 2012, Medical Electrical Equipment - Part 1: General Requirements for Basic Safety and Essential Performance
- IEC 60601-1-8 Ed. 2.1: 2012, Medical Electrical Equipment - Part 1-8: Collateral Standard: Tests and Guidance for Alarm Systems in Medical Electrical Equipment and Medical Electrical Systems.
- IEC 60601-1-11 Ed. 2.0: 2015, Medical Electrical Equipment - Part 1-11: Collateral Standard: Requirements for Medical Electrical Equipment and Medical Electrical Systems Used in the Home Healthcare Environment
- ISO 23908: 2011, Sharps injury protection – Requirements and test methods – Sharps protection features for single- use hypodermic needles, introducers for catheters and needles used for blood sampling
- IEC 62304 Ed. 1.1: 2015, Medical Devices Software - (Software life cycle processes)
VI Performance Characteristics:
A. Analytical Performance
1. Basal delivery accuracy
To assess basal delivery accuracy, 12 Omnipod GO pumps were tested by delivering at low, medium, and high basal rates (10, 25, and 40 U/day). Water was used as a substitute for insulin. The water was pumped into a container on a scale and the weight of the liquid at various time points was used to assess basal delivery accuracy. The following tables report the typical basal performance (median) observed, along with the lowest and highest results observed for the low, medium, and high basal rate settings for all pumps tested with no warmup period. For each time period, the tables show the volume of insulin requested (expected delivery volume), the volume that was delivered as measured by the scale, and the minimum and maximum delivery errors as a percentage of the expected delivery volume.
K223372 - Page 5 of 9
{5}
Table 1: Amount of fluid delivered after 1, 6, and 12 hours with ${10}\mathrm{U}/\mathrm{{day}}$ (low) basal rate setting
| 10 U/day Basal Duration | 1 hour | 6 hours | 12 hours |
| --- | --- | --- | --- |
| Total expected delivery volume | 0.42 U | 2.50 U | 5.00 U |
| Median amount delivered [min, max] | 0.42 U [0.14, 0.64] | 2.53 U [1.72, 3.09] | 5.05 U [3.83, 5.18] |
| [min, max] Error % | [-67%, 55%] | [-31%, 24%] | [-23%, 3.6%] |
Table 2: Amount of fluid delivered after 1, 6, and 12 hours with ${25}\mathrm{U}/\mathrm{{day}}$ (medium) basal rate setting
| 25 U/day Basal Duration | 1 hour | 6 hours | 12 hours |
| --- | --- | --- | --- |
| Total expected delivery volume | 1.04 U | 6.25 U | 12.50 U |
| Median amount delivered [min, max] | 1.07 U [0.46, 1.54] | 6.34 U [5.09, 6.88] | 12.59 U [11.31, 13.39] |
| [min, max] Error % | [-56%, 47%] | [-18%, 10%] | [-9.5%, 7.1%] |
Table 3: Amount of fluid delivered after 1, 6, and 12 hours with ${40}\mathrm{U}/\mathrm{{day}}$ (high) basal rate setting
| 40 U/day Basal Duration | 1 hour | 6 hours | 12 hours |
| --- | --- | --- | --- |
| Total expected delivery volume | 1.67 U | 10.00 U | 20.00 U |
| Median amount delivered [min, max] | 1.73 U [0.99, 2.28] | 10.09 U [9.09, 10.37] | 20.19 U [19.13, 20.47] |
| [min, max] Error % | [-41%, 37%] | [-9.1%, 3.7%] | [-4.3%, 2.4%] |
# 2. Occlusion detection
An occlusion hazard alarm sounds when an average of 3 units to 5 units of missed insulin occurs. A total of 21 pods were tested for the time to occlusion alarm at the applicable basal rates (10, 15, 20, 25, 30, 35, and $40\mathrm{U/day}$ ) by introducing occlusion to the pods, recording the time for occlusion, and parsing all alarm codes related to occlusion to record occlusion alarm time. The time to occlusion alarm is the difference between the time of occlusion and the time of occlusion alarm.
The following table depicts blockage detection for three different situations when using U-100 insulin. For example, if the pod's cannula becomes blocked while using a 25 U/day pod, up to 5.5 hours may pass before the pod sounds a hazard alarm.
K223372 - Page 6 of 9
{6}
Table 4: Timing of occlusion detection alarms
| Basal | Typical time to occlusion detection, hours | Maximum time to occlusion detection, hours |
| --- | --- | --- |
| 10 U/day | 8.5 | 12.5 |
| 25 U/day | 3.5 | 5.5 |
| 40 U/day | 2.5 | 3.5 |
## B. Other Supportive Instrument Performance Characteristics Data
### 1. Hazard Analysis
The subject device utilizes an insulin pump design that is similar to that of the Omnipod DASH Insulin Management System cleared in K211575. A comprehensive hazard analysis was reviewed in K211575, in which design inputs and outputs, risks, and risk mitigations for hardware and software associated with proper functioning of the insulin pump were reviewed.
As part of this submission, the sponsor provided an updated hazard analysis to account for the unique intended use, design elements, and risks of the Omnipod GO device that were not previously reviewed. This analysis identified hazards which could reasonably be anticipated to impact the proper use of the device, traced all identified risks to adequate design controls, and demonstrated that design features were appropriately implemented and validated.
### 2. Human Factors
Human factors validation tests were conducted with the Omnipod GO Insulin Delivery Device. All study participants received training that was consistent with the training that patients would receive with the commercial product. Usability evaluations assessed comprehension and usability of the device for critical device tasks. Results of the study demonstrated that the Omnipod GO Insulin Delivery Device is validated for its intended use.
### 3. Biocompatibility
All of the materials and manufacturing processes of the proposed Omnipod GO pod patient-contacting and fluid path-contacting components are identical to those of the Omnipod DASH pod (K211575). The differences between the GO and the Omnipod DASH pods are associated with the printed circuit board assembly (PCBA), which is non-patient and non-fluid path-contacting. Therefore, biocompatibility testing previously performed on the fluid path and patient-contacting components of previously cleared versions of this device were leveraged.
Additional testing including in vitro cytotoxicity of the device in its final package, and off-gassing of the GO pod PCBA were performed to ensure product quality and design consistency. The biocompatibility evaluation of the GO Pod PCBAs via off-gas analysis
K223372 - Page 7 of 9
{7}
show no detectable chemical or toxicological concerns and the cytotoxicity study of GO Pods for fluid path components and adhesive pads showed no evidence of in vitro cytotoxicity.
## 4. Sterility
An EO sterilization product adoption was completed in accordance with AAMI TIR 28:2016. The same sterilization validation method described for the Omnipod DASH device (K211575) continues to be applicable to the subject device. The sterility information was reviewed and found to be acceptable.
## 5. Insulin Compatibility and Stability
Omnipod GO Pod is compatible with U-100 insulins Novolog, Humalog, Admelog, Fiasp, and Lyumjev just like the Omnipod DASH and Eros systems. No additional insulin compatibility testing was conducted with Omnipod GO Pods as the insulin contacting portions of the Pod is identical to DASH and Eros.
## 6. Additional Bench Testing
The subject device is evolved from the currently marketed system found in the Omnipod family of Pods including the Omnipod DASH device (K211575) and shares many aspects in product design and requirements. Select verification information of previously cleared Omnipod devices was leveraged for the subject device. Additional verification testing that was conducted specifically for the subject device is as follows:
| Shelf Life Test |
| --- |
| Battery Life Test |
| Start Up Conditions Test |
| Infusion Pressure Test |
| General Features and Functions Test |
| Pod Activation, Deactivation and Insulin Delivery Initiation and Termination Test |
| Environmental Operating Range Test |
| User Interface Test |
| Occlusion Release Test |
## 7. Electromagnetic Compatibility and Wireless Coexistence
EMC of the subject device has been evaluated per the following standards:
- IEC 60601-1-2:2020 (ed. 4.1), Medical Electrical Equipment, Part 2: Collateral standard: Electromagnetic compatibility requirements & test
- RTCA DO-160G: 2010 Environmental conditions and test procedures for airborne equipment
- IEC 60601-2-24 Edition 2.0, 2012-10, Medical electrical equipment – Part 2-24: Particular requirements for the basic safety and essential performance of infusion pumps and controllers o Essential performance defined in this particular standard
K223372 - Page 8 of 9
{8}
- IEC 60601-1-11:2015 +A1:2020, General requirements for basic safety and essential performance – Collateral Standard: Requirements for medical electrical equipment and medical electrical systems used in the home healthcare environment
All applicable emissions and immunity tests for battery-powered ME equipment were conducted per IEC 60601-1-2:2020 and RTCA DO-160G to demonstrate that the device will perform as expected in the home healthcare environment.
8. Basic Safety and Essential Performance (Electrical Safety)
Insulet provided verification evidence for compliance with the IEC 60601-1 and applicable collateral standards. Verification results support the finding of substantial equivalence for this device.
9. Software
Detailed information on software of the device was reviewed and found to be acceptable.
10. Cybersecurity
Detailed information on Cybersecurity of the device was reviewed and found to be acceptable.
VII Proposed Labeling:
The labeling supports the finding of substantial equivalence for this device.
VIII Conclusion:
The submitted information in this premarket notification is complete and supports a substantial equivalence decision.
K223372 - Page 9 of 9
Predicate graph will load when search results are available.
Embedding visualization will load when search results are available.
PDF viewer will load when search results are available.
Loading panels...
Select an item from Submissions
Click any panel, subpart, regulation, product code, or device to see details here.
Section Matches
Results will appear here.
Product Code Matches
Results will appear here.
Special Control Matches
Results will appear here.
Loading collections...
Loading
My Alerts
You will receive email notifications based on the filters and frequency you set for each alert.
Sort by:
Create Alert
Search Filters
Agent Token
Create a read-only bearer token for Claude, ChatGPT, or other agents that can call HTTP APIs.
Copy this now. It will not be shown again.
Connected apps
Apps you authorized through browser sign-in. Disconnecting revokes their access immediately.
Learn the FDA Browser
Two short videos show you everything — or skip straight to the written tutorial if you'd rather read. You can reopen this any time from the Tutorial button in the top bar.
Part 1 — Search, results, and everyday workflows 16 min
Part 2 — Embeddings: the galaxy map 3 min
1. Search: exact and fuzzy
Type a phrase like "coronary artery calcification" into the search box. You get two kinds of results. Exact results match the literal phrase — prefix searches work ("coronary artery calcificati") but suffix searches do not. Fuzzy results match on the meaning and intent of your phrase rather than the exact words, and are sorted by relevance score. Hover over the Exact or Fuzzy badge on any row to see exactly why it matched.
Use the checkboxes above the results to narrow: SaMD keeps only software-only devices, AI / ML keeps only devices with AI.
Exact vs. fuzzy search: what's the difference?
Exact matches on the literal phrase (prefix search works, suffix does not). Fuzzy matches on the meaning and intent of the phrase rather than the exact words. Hover over the badge on any row to see why it matched.
You search "coronary artery calcification" and want only software devices with AI. What two filters do you apply?
Narrow by SaMD (software-only devices), then narrow by AI/ML (devices with AI).
2. The results table
Scroll right in the results table. The intended use is extracted for you — no need to open the PDF. The device story gives a high-level snapshot of what the device does and how it's used. The AI Performance sub-table shows each output name, acceptance criteria, observed values, and development/test dataset descriptions — the same format Innolitics uses for regulatory strategy outputs, and the fastest high-level fingerprint of an AI device. It is AI-generated but has been very reliable in practice.
Where do you find a device's intended use without opening the PDF?
Scroll right in the search results table. The intended use column is extracted for you; no need to dig into the 510(k) summary PDF.
What does the AI Performance sub-table show, and why is it useful?
Output name, acceptance criteria, observed values, development dataset description, and test dataset description. It's the same format we use for regulatory strategy output and Fast 510(k) input, and the fastest high-level fingerprint of an AI device. AI-generated but reliable in practice.
3. Judging fuzzy relevance
Fuzzy results trail off in relevance as you scroll. Use three signals to decide how far down to go: the fuzzy badge explanations, the intended use column, and whether your target output (e.g., Cobb angle) still appears in the AI Performance sub-table. Once it stops appearing, you're past the relevant zone. A top hit with a low score (~0.4) and a stretched explanation is a hint the closest predicates are far away — the project may be headed for De Novo. Note the fuzzy search is a pattern match: it doesn't handle negation ("not") well, and hardware devices can appear — filter by SaMD/AI ML to cut them.
How do you judge how far down fuzzy search results to go?
Use the relevancy signals: the fuzzy badge explanations, the intended use column, and whether the target output (e.g., Cobb angle) still appears in the AI Performance sub-table. Once it stops appearing, results are trailing off in relevancy.
4. Device detail page: chat and citations
Click a device name to open its detail page: device facts on the left, a chat window on the right. Ask something like "Describe the training data". The answer carries little citation bubbles — click one to jump to the highlighted passage in the source PDF, so you can verify every AI answer against the document. There's also a Download PDF button for sharing.
How do you verify an AI chat answer on the device detail page?
Click the citation bubbles to jump to the relevant highlight in the source document.
Reading rule for every project: how many summaries do you read in full?
At least the three most relevant 510(k) or De Novo summaries, in full. After that, use targeted chat questions to confirm your memory quickly. The tool supports this professional habit — it doesn't replace it.
5. Side-by-side comparison
Select multiple rows in the results table (aim for under ~10), then open the PDF Viewer tab. Ask one question — it goes to all selected devices in parallel, each with citations. This is the fastest way to compare and contrast devices: training data, PCCP scope, how they handled adding new scanners, and so on.
What does the side-by-side PDF viewer mode do?
Select multiple devices, open the PDF viewer tab, and ask one question (e.g., "Describe the training data"). It queries all selected devices simultaneously with citations, so you can compare and contrast quickly.
6. Collections
With rows selected, go to the Collections tab and create a labeled collection (e.g., "Cobb Angle Project"). Reload that selection any time — before a client call, pull up the collection and ask questions across all of its devices at once.
How do you save a set of selected devices for later use?
Select the rows, go to the Collections tab, and create a labeled collection (e.g., "Cobb Angle Project"). You can reload the selection anytime and carry it into the PDF viewer and other tabs that support selections.
7. Product codes and the regulations tree
Click a product code in the results to jump to it in the regulations tree — identification text, sibling product codes, and devices you can open in a PDF viewer on the right. Click a regulation number to see its identification, special controls, and related product codes. You can also search by product code or regulation number at the top of the tree. Always read the special controls if any exist for your device — it broadens your search and sharpens pre-kickoff research.
What can you do from the regulations tree view?
Browse product codes and regulation numbers, read the identification text and special controls, browse sibling product codes, open device PDFs on the right, and search by product code or regulation number at the top of the tree.
8. Chart view
Click Show Chart and segment by regulation number (or product code) to see which regulations dominate your result set. Clicking a regulation takes you into the regulations tree. Great for spotting that most matches are, say, hardware laparoscopic devices — a cue to go back and filter.
How do you see which regulations dominate a search result set?
Click "Show Chart" and segment by Regulation Number. Clicking a regulation takes you to the regulations tree.
9. The predicate graph
Open the Predicates tab for a family-tree view of predicate relationships. Click a node to trace its parents and children; selections from search carry over pre-selected. Commonly predicated devices are worth reading — a lot of people predicated them for a reason. The visual lineage is also handy on client calls, e.g. to show how a predicate family evolved and justify why your predicate still holds.
In the predicate graph, why are commonly predicated devices worth reading?
A lot of people predicated them for a reason. Clicking a node traces parents and children, and selections from search carry over pre-selected.
10. Embeddings: the galaxy map
The Embeddings tab plots every matching document in a 2-D "galaxy map" where semantically similar devices cluster together. Hover or click clusters to explore, and let AI label the clusters for you. Embeddings beat product codes for grouping: two devices can carry different product codes (LLZ vs. QIH) yet do the same thing — the embedding captures the meaning of the intended use and device story. This is also exactly how retrieval-augmented generation (RAG) works under the hood, and it makes a great visual on client calls.
Try it yourself
Head to the search page and work through a few of these AI/ML fuzzy searches to build intuition: perivascular fat on CT · aortic valve calcification opportunistic screening on noncontrast CT · breast cancer prediction on digital pathology slides · autism detection · gestational age prediction · a hearing aid that can also detect a pulse · foundation model based analysis of ECG · large language models · penetration test. Watch how the relevance scores, intended use, and AI Performance tables tell you when results stop being meaningful.