Monitor Management and Review Workflows
Once a monitor is running, use the dashboard to track progress, manage assignments, and refine your review process over time.
Monitor dashboard
Access the dashboard from the main menu or at app.amplified.ai/monitors . Each monitor shows:
- Name and description
- Required tags for review
- Last updated date and next scheduled update
- Workspace scope and assigned members
Only the monitor's owner can edit, pause, or delete the monitor. Admins and the owner can view monitor details.
View assignment history
Click View assignments on any monitor to see all past and current assignment cycles. Each entry shows:
- The date the assignment was generated
- Review progress — how many patents have been reviewed vs. total
- Review type — what criteria were assigned (simple relevance, tag classification, etc.)
- Assigned reviewers — who was asked to review
This gives you a complete history of your monitoring activities and their completion status.
Add reviewers to an existing monitor
To add a new team member to a monitor that's already running:
- Open the Monitor dashboard and click the ... menu next to the monitor
- Click Edit
- Click Next to reach the Assign Review step
- Add users from the Assign to section
- Save
New reviewers will be included starting from the next cycle. They won't retroactively receive previous assignments.
Note: The user must already be a member of the monitor's workspace. If they're not, add them to the workspace first.
Track review progress
Each assignment tracks individual reviewer progress. You can see at a glance:
- Which patents have been reviewed (green icons)
- Which patents still need review (red/orange icons)
- Overall completion percentage
If a review assignment has multiple users, their review syncs automatically. If one reviewer logs in and reviews 10 results, the next reviewer will see those as already reviewed. There's no risk of duplicate work — everyone sees the same progress in real time.
When all reviews in a cycle are complete, the monitor owner receives a notification (if enabled).
Multi-step review workflows
Some monitoring tasks need more than one round of review. Amplified supports this through sequential assignments:
Step 1 — Initial triage: First-round reviewers classify results for general relevance.
Step 2 — Specialist review: Assign relevant results to domain experts for detailed analysis.
Step 3 — Legal review: Assign critical findings to legal for triage and action planning.
To set up additional steps, open the monitor's project, select the results that passed the previous review, and create a new assignment with different reviewers and instructions.
Each step has its own progress tracking and notifications.
Classifier improvement over time
When your monitor uses tag-based review with classifiers:
- Each cycle, the classifier predicts tags on new results
- Reviewers confirm or reject predictions
- After the cycle, classifiers are automatically retrained on the updated data
- Next cycle, predictions are more accurate
This creates a feedback loop that gets better with every cycle. After several rounds, the classifier handles the majority of obvious cases, and reviewers focus only on borderline patents.
De-duplication
By default, monitors show new families — if a family member was already reviewed in a previous cycle, the family won't appear again. This prevents duplicate work.
Switch from families to publications if you need to see every new document, even from previously reviewed families (e.g., when tracking the status of specific applications).
Pausing and editing monitors
Pause a monitor to temporarily stop generating new assignments without losing your history. Resume at any time.
Edit a monitor to change the query, frequency, reviewers, or tags. Changes take effect from the next scheduled run.
Delete a monitor to permanently remove it. Assignment history is preserved in the monitor's project.
Related articles
→ Set up automated patent monitoring — creating a new monitor
→ Reviewing monitor results — reviewer's guide to completing assignments
→ Train an AI classifier — improving classifier accuracy
→ Assign patents for review — creating ad hoc assignments outside of monitors