Frequently Asked Questions

TrainBase is a training platform designed for medical practices. Doctors and staff can record screen workflows or mobile camera videos, which are then processed to create professional training videos with automated narration, assisted PHI flagging, and quizzes. It turns "watch me do this" into polished, reusable training content.

Yes. You can book 15 minutes with Dr. Szender, a practicing gynecologic oncologist who built TrainBase for his own clinic. He will walk a recorded workflow through to a finished training video on a sample practice chart, and tell you plainly whether it fits a practice like yours. Please do not share live patient records on the call — the demonstration runs entirely on sample data, so there is no need.

Both. Your practice's own recordings stay private to your practice — they are never visible to other organizations unless you explicitly share them. On top of that, TrainBase Official videos are finished training videos produced by the TrainBase team and made available to every organization on the platform, so your team can start learning from professionally produced content right away. They appear in the Training Room and in search results tagged "TrainBase Official" and play just like your own organization's videos. They're view-only — anyone can watch them and take their quizzes, but only the TrainBase team can edit or remove them.

TrainBase is built for medical practices of all sizes — from solo private practices to multi-location groups. It's used by:

• Doctors & clinical staff who want to record workflows as training material
• Office managers & admins who need to onboard and train new hires efficiently
• New employees & trainees who need to learn procedures quickly

TrainBase provides tools to support HIPAA compliance, including automated PHI detection (using OCR and entity recognition), manual blur/redaction tools, and a Business Associate Agreement (BAA). Real-time redaction during browser-based screen recording runs entirely in your browser using same-origin libraries. A server-side redaction agent also reviews sampled video frames to catch additional PHI before publishing; it is on by default for every workspace, and those frames are processed by Google Cloud under our BAA and are never used to train any model. Admins can turn it off at any time. Compliance is a shared responsibility — users should avoid recording patient-identifiable information whenever possible and always review videos for PHI before publishing.

TrainBase works best in Google Chrome or Microsoft Edge on desktop for recording. Safari has limited screen recording support. For mobile uploads, any modern iOS or Android browser works. Video playback and training modules work on all modern browsers and devices.

Yes — a Chrome and Edge extension lets you search your practice's training library from any tab without having to open TrainBase. A small search pill sits at the edge of every page; type a question and answers plus matching videos open in a side panel. To install, sign in to TrainBase, go to your Profile, and click "Install the TrainBase browser extension", then "Add to Chrome" — it's a one-click install from the Chrome Web Store and updates itself automatically. IT admins who'd rather deploy it by hand can still download the .zip from the same page and load it unpacked.

Only the search text you type into its own pill. The extension never reads the content of the page you're on, never sees your patient charts, and never sees your EHR. Sign-in uses a short-lived token (default 8 hours, one clinical shift) that you or your admin can revoke instantly. Signing out of TrainBase on the web also revokes any active extension sessions for your account.

Yes, if your admin turns it on. An admin can issue a reusable access key for a trainee from Admin → Trainee Sign-in Settings. The trainee adds that key to their extension once and can then search and watch without a sign-in screen. Each key belongs to one trainee, can be turned off at any time, and can be replaced if it's ever exposed — so it's never permanent.

Visit the registration page and enter your name, email, and a password. If you've been invited to an organization, use the invite link your admin sent you instead — it will automatically connect you to the right workspace.

Visit our Pricing page for current plan details. Plans are billed monthly or annually per workspace, not per user. Trainees who join via invite are covered by the organization's subscription at no extra cost. Every new account includes a 14-day free trial with full access to all features — no credit card required to get started.

The workspace Owner can cancel the subscription from the Admin panel by clicking "Manage Billing", which opens the Stripe customer portal. From there you can cancel, pause, or change your plan. Your access continues until the end of the current billing period.

Ask your admin for an invite link. If your organization has enabled auto-join, simply create an account with your company email address, verify your email, and you'll be placed in the correct workspace automatically. If multiple practices share your email domain, you'll be asked to select which practice you belong to — your admin will then receive an email to confirm your access before you're added. If you're already registered but not in a workspace, contact your admin to receive an invite.

TrainBase uses a multi-layered approach to PHI protection. Videos are automatically scanned using OCR (optical character recognition) combined with automated entity analysis to detect names, dates, medical record numbers, and other sensitive information. Detected regions are highlighted for review, and you confirm or adjust them in the editor before publishing. All flagged areas are rendered as a pixelated mosaic in the published video — the same effect as the live recording assist — ensuring PHI never appears in training content.

Your plan includes a set number of processed videos each month (20 on Standard, 100 on Growth, unlimited on Enterprise). A video counts when it is processed — that is, when you record it, upload it, create a clip from an existing video, or start a brand-new draft — not when you publish it. Recordings that fail to process do not count. (If one is interrupted by maintenance and still shows against your total, contact us and we’ll credit it back.) When you reach the limit, recording and uploading pause until the allowance resets at the start of the next month; everything already in your workbench can still be edited, published and watched. The recording page shows how many days remain.

Agentic redaction is on by default for every workspace, including free trials. It adds a vision model that looks at each sampled frame and reasons about what is on screen — so it recognizes a patient name in a chart header, reads handwriting on a held-up document, and spots visible faces — and runs alongside the standard scan. It processes frames on HIPAA BAA-covered cloud infrastructure, and it is included in your plan — there is no per-video or per-minute charge for it. Admins who prefer to disable it can do so any time under Admin > User Management > Trainee Sign-in > Agentic redaction policy, and that choice is always respected.

Where the redaction agent is enabled, every publish — including videos published with no manual edits at all, and videos approved from the admin review queue — runs a final verification pass on the finished video itself: TrainBase re-scans the output at moments in between the original detection samples — including where content was scrolling fastest — and confirms nothing identifying slipped through, whether that's readable text or a visible patient face. Anything still identifiable is covered automatically before the video is released, and you're notified when that happens. If verification cannot run or cannot confirm the video is clean, the video is held back rather than published, and you can simply try publishing again.

In every case, the review step in the editor is the authoritative pass: you confirm and adjust the detected regions, and nothing publishes until you do. If redaction fails for any reason, the video is held in a failed state rather than published.

Not unless you explicitly let them, and only briefly. If you need hands-on help, a workspace admin can grant TrainBase support temporary access from Admin > User Management > TrainBase Support access, choosing a duration of 5, 15, 30, or 60 minutes. During that window, support sees your workspace the way an admin does — but cannot delete anything, change billing or sign-in settings, invite users, or view raw (unredacted) recordings. Access ends automatically at the time you chose, the moment support signs out, or the instant you click Revoke now — whichever comes first. Every support entry and every action taken during the session is recorded in your workspace audit log.

Yes. Passwords are hashed using industry-standard bcrypt encryption and are never stored in plain text. During registration, your password is checked against the HaveIBeenPwned database of known breaches to ensure it hasn't been compromised. Passwords must also meet complexity requirements including minimum length and character variety.

TrainBase uses three categories of cookies:

• Essential (required): These keep you logged in and enable core functionality. They cannot be disabled.
• Analytics (optional): These help us understand usage patterns and improve the platform. No personal data is shared with third parties.
• Preferences (optional): These remember your settings like theme and layout choices.

You can manage your cookie preferences at any time by clicking "Cookie Settings" in the footer.

Go to Profile > Security and click "Enable MFA". You'll need an authenticator app on your phone — TrainBase works with any TOTP-compatible app, including:

• Google Authenticator (iOS · Android)
• Microsoft Authenticator (iOS · Android)
• Okta Verify (iOS · Android)
• Authy (iOS · Android · Desktop)

Once you have an app installed, scan the QR code shown on screen (or enter the manual key) to register your account. Enter the 6-digit code from your app to verify, and you're set. MFA is required for all users. You'll also receive backup codes — save them in a safe place in case you lose access to your authenticator app.

Go to Profile > Account where you'll find two options:

• Export My Data: Downloads a copy of all your personal data, watch history, and quiz scores.
• Delete Account: Permanently removes your account and associated data. Deletion includes a 30-day grace period during which you can cancel the request by logging back in.

By default, deleted videos are retained for 90 days before permanent removal, giving you time to recover accidentally deleted content. Workspace owners can customize this retention period (from 30 days up to 7 years) in Admin > Workspace Settings. Administrators can also export a complete audit log of all activity in their workspace as CSV or JSON from the PHI Audit Log tab. Unredacted original source recordings are held only transiently: an original that is never linked to a saved video is permanently deleted within about a day, and an original linked to a saved training video is permanently deleted within 30 days — after which only the redacted version is kept. Our Business Associate Agreement sets out the full raw-recording retention and destruction schedule.

Click the Log Out button in the navigation menu or on your Profile page. You will be logged out immediately. For your security, sessions also expire automatically after 15 minutes of inactivity.

Go to the Record page (Bounty Board) and click "Record Screen". Your browser will ask you to select a screen, window, or tab to share. Perform your workflow, then click "Stop Recording" when finished. The video will load into the editor for editing and submission. See the User Manual for detailed steps.

While you're recording your screen, TrainBase can pixelate faces and on-screen text directly in your browser, as you record. The file that gets uploaded is the already-blurred version — no raw-with-PHI copy is sent to our servers and blurred later. A banner at the top of the recorder shows live counts: "Blurring 1 face(s), 4 text region(s)". The workbench still does a full review after upload, so you get two layers of protection. Real-time privacy blur is always on wherever your browser supports it (Chrome and Edge); on Firefox and Safari the workbench review covers it instead. The workbench review runs either way. See the User Manual for details.

Recording is never blocked by a blur failure. If the detectors can't start (unsupported browser, slow network, or the initialization throws), the banner changes to "Real-time blur unavailable — your recording will go to the workbench for full review" and the recording continues with the raw stream. If the blur stops working partway through a recording, the banner switches to the fallback message within two seconds and the same behavior applies. Any recording that didn't get full real-time blur shows a "Needs full review" badge on its workbench card and sorts to the top of your active drafts so you see it first.

Chrome and Edge (version 120 or later) support the full real-time blur pipeline. Firefox and Safari don't yet expose the underlying browser APIs we need, so they fall back automatically — recording still works, the banner tells you blur isn't running, and the workbench does the full review after upload. No action needed on your part; just know that you'll see the "Needs full review" badge on those recordings in your workbench.

On the Record page, tap "Upload from Phone" and select a video from your camera roll. The video is uploaded and processed automatically. You can also use the file upload button to choose a pre-recorded file from your computer.

TrainBase accepts MP4, MOV, WebM, and AVI files. All uploads are automatically converted to MP4 for processing. For best results, aim for 720p or 1080p resolution.

Each account can upload up to 5 GB of video per day. This limit resets automatically at midnight UTC. If you need to upload more in a single day, wait until the limit resets or contact your workspace admin. Large screen recordings can be reduced in size by trimming the video before uploading or lowering your screen resolution before recording.

If screen recording isn't starting:

• Make sure you're using Chrome or Edge (Safari has limited support)
• Check that you've granted the browser permission to record your screen
• Try refreshing the page
• See our Camera Help guide for detailed troubleshooting

In the video editor, select the "Blur" tool and draw a rectangle over the area you want to redact. Click the blur zone to select it, then use the "Start Here" and "End Here" buttons (or press Space) to set when the blur should appear and disappear. You can add multiple blur zones to a single video. In the published video, each zone is rendered as a pixelated mosaic. Press Esc to deactivate the tool and return to normal pointer mode. Your redaction zones are saved with the draft, so if you re-edit the video later, your blur zones will still be there. Note that re-editing requires the original source recording, which is kept for a limited time (up to 30 days after the video is saved — see the data retention question below). Once the original has been removed, the video can no longer be re-edited or re-scanned; the published version itself is unaffected. A draft you never published is affected differently: its source recording is the only copy, so once the 30-day window passes the draft can no longer be played or edited at all. Publish a draft you want to keep, well before it reaches 30 days.

TrainBase renders redacted areas as a pixelated mosaic whenever possible. If that effect can't be produced for a particular video, it falls back to a blur, and only if that also fails does it use solid boxes. You'll see a notification after processing telling you which was used. All three cover the area completely — the difference is cosmetic, and the video is never released with less coverage.

Nothing was published — that is what the failed state means. The verification pass found something identifying it could not cover on its own, so TrainBase held the video back instead of releasing it.

You have two ways forward, and you can use both. Re-run Redaction — in the workbench draft editor, or from the video's menu in Admin > Content — scans the video again from scratch, re-renders it and re-verifies it; use this when you think the scan simply missed something. If it fails a second time, open the editor and draw blur boxes over the remaining areas by hand: your manual boxes are always applied, and the review step in the editor is the authoritative pass.

While a video is unpublished — whether it failed, is still a draft, or you de-published it — its original recording can only be opened by you and by the people who work on videos in your practice (owners, admins and content creators). Trainees cannot open it.

Crop: Select the "Crop" tool and draw a rectangle to define the visible area. Only the cropped region is included in the final video.

Trim: Drag the left and right handles on the timeline bar to cut the beginning or end of your video. The trimmed portions will be removed from the final output.

When you submit a video, TrainBase analyzes the video content and generates a timestamped narration script. This script explains each step of the workflow in clear, professional language. The script is then converted to a voiceover and overlaid on the final video. You can edit the script before it's finalized.

Yes. After the script is generated, you can open the video in your Workbench draft editor and modify the script text. You can also use the "Refine Narration" button to adjust tone, add detail, or restructure the narration.

Yes. In the draft editor, the Quiz Questions section shows each generated question with editable fields for the question text, four answer options (A–D), the correct answer, and an optional explanation. Add or remove questions with the controls in each card. Changes save when you click Save Draft.

Processing time varies depending on video length and complexity. You'll see a real-time progress bar showing each stage of the pipeline: PHI analysis, redaction, narration generation, and final assembly. Most videos under 5 minutes complete within a few minutes.

Detection runs at a fixed, deliberately cautious level — there is no dial to set. The redaction agent reads each sampled frame and the standard text scan runs behind it, so borderline items are flagged rather than missed. Because that errs toward over-flagging, you review every proposed blur box in the editor and can remove any that cover non-sensitive content before publishing.

If video processing encounters an error, the video is marked with an "Error" status and you'll be notified. You can retry processing from the Workbench. Processing automatically times out after 30 minutes to prevent stuck jobs. If the problem persists, try re-uploading the video or reducing its resolution.

No. As soon as you stop a screen recording, TrainBase saves it and starts the privacy scan on our side, and that scan keeps running even if your tab closes or the page reloads. Open your Workbench and look under Unfinished Recordings: click Review & Edit if the scan has finished, Reconnect if it is still running, or Run the scan again if it was interrupted (it starts over from the beginning). While a scan is running you'll see what it's doing and, during the detailed pass, how many seconds have been reviewed, so a long recording is never mistaken for a stuck one. An unfinished recording is kept for 24 hours, then deleted, so please come back to it within a day or record it again. Very close to that deadline you can still open a finished scan, but starting a fresh one is turned off — there wouldn't be time for it to finish. Until you submit it, an unfinished recording is visible only to you: nobody else in your practice can open it, and no one can delete it while its scan is running.

No. When you stop a screen recording, TrainBase looks at what was on screen and, if it sees more than one distinct workflow, offers to split the recording into separate training videos before anything else happens. You'll see "N workflows detected in this recording" with a preview and a timeline.

Nothing is decided for you. Drag the green markers to move a section break, click Add section break at playhead to create one, rename any section, merge ones that belong together, and click the trash can on any section you don't need — then either Split into N videos or Keep as one video. Each section you keep becomes its own draft with its own narration, its own quiz and its own privacy scan, so a new hire can be assigned just the one task they need instead of a twenty-minute recording with their part buried in the middle.

Drop what you don't need before you split, not after. A dropped section is never cut into a video, never scanned, and never counts against your allowance — so a recording where only two of five sections are worth keeping costs you two videos, not five. Dropping is reversible right up until you press Split: the row dims, its span goes grey on the timeline so you can see what is being left out, and an arrow puts it back. A split needs at least two sections, so the trash cans grey out once two are left.

Two practical notes. The preview at this stage is unredacted — it is your own recording, visible only to you, and the privacy scan runs on each section after you split. And each section counts as one video against your monthly processed-video allowance, while the original recording's own count is released — so splitting a recording into three uses three, not four. If your remaining allowance can't cover the split, TrainBase tells you rather than doing it halfway.

Because scanning is the slow part, and scanning footage you were going to throw away is wasted time. The section proposal is a quick look at what's on screen, not a privacy scan — it costs seconds. The detailed scan then runs on each section you actually keep, in parallel, so you wait once for the material you want instead of once for everything you recorded. Each proposed section carries a rough PHI density badge (low, medium or high) as a heads-up about which parts will need the most attention from you in the editor.

The Bounty Board is a list of training videos your organization needs. Admins create bounties (often generated from SOP documents), and hunters claim and record them. Each bounty includes a title, difficulty level, point value, and guidance notes (visual goal, director's note, script prompt).

You earn points when a video you created is approved and published by an admin. The points awarded match the bounty's difficulty:

• Easy: 50 points
• Medium: 100 points
• Hard: 200 points

If a video doesn't match a specific bounty, you still earn 100 points as a default. Your total bounty points are displayed on your Profile page.

On the Bounty Board, click any bounty to expand its details. Click "Start This Bounty" to begin — the bounty name will be loaded into the recording form. Record or upload your video, then submit it for review. Once your admin approves and publishes it, you earn the bounty points and the bounty is marked as completed.

Click "Train" in the navigation bar. Any modules your admin has assigned will appear at the top. Click a module to see its lessons, then click a lesson to start watching. Your progress is automatically saved.

Quizzes are scored on a 0-100 scale. Your score is calculated as a percentage of correct answers. Scores are visible to you and your admin. You can retake quizzes to improve your score — your highest score is kept.

Quizzes are off by default for every workspace. A workspace admin can enable them under Admin → Workspace Settings → Quizzes. Once enabled, any new video submitted will include an auto-generated multiple-choice question. Videos published before the toggle was turned on will not have a quiz retroactively unless the admin re-generates the automated script for that video.

Yes. All videos remain accessible in the Training Room after completion. You can also find previously watched videos in your Profile's Watch History tab or in your Favorites if you've bookmarked them.

Admins and hunters can assign specific videos to trainees from the training management section. Select the videos you want to assign, choose the team members, and optionally set a due date. Assigned trainees will see the training appear at the top of their Train page with the due date displayed.

The Training Leaderboard shows the top content creators in your organization, ranked by training impact. Rankings are based on a combination of video views, training completions, and quiz scores achieved by trainees who watched the creator's content. It's a great way to recognize team members who contribute the most valuable training material.

Use the Training Mentor search bar at the top of the Train page. Type what you're looking for and the mentor will suggest the best matching video with relevant timestamps. If the suggestion isn't what you need, click "Not what I need" to see all other matches and a "Request New Video" option.

On Clinic Growth and Enterprise plans, the Training Mentor also answers your question in text, drawn only from your practice's own training video scripts — it never uses outside knowledge or makes up an answer. Each answer shows a "From your training library" line citing the source video(s), with clickable chips that jump straight to the relevant moment in the video. The matching video suggestions still appear below the answer.

The mentor tells you honestly that your practice's library doesn't cover it — it will never guess. You'll see the closest matching videos, plus a "Request a Video on This" button that adds the topic to your team's recording queue. Your practice's admins are also notified automatically: the question appears in their Content Gaps list so they know what training content is missing.

Yes. When your question describes a task that spans several videos (for example, "walk me through new hire onboarding from start to finish"), the mentor detects that the task has multiple steps and offers to generate a playlist of the matching videos, ordered from task start to task completion. Click "Yes, create playlist" to save it to your playlist dropdown, or "Queue in order" to load the videos directly into your training queue.

Yes. If the Training Mentor can't find what you need, click "Not what I need" on the suggestion, then use the "Request New Video" form at the bottom. Describe the training you need and it will be added to the recording queue for your team's content creators.

Go to the Admin panel and find the video in the review queue. Click on it to preview the video and script. Use "Approve & Publish" to generate the voiceover and publish, or "Reject" with feedback to send it back to the hunter for revision.

In the Admin panel, go to the Team section and click "Invite Member". You can generate a shareable invite link or send an email invitation directly. You can also enable auto-join for your company's email domain so anyone with a matching email is automatically added when they register.

In the Admin panel, open the User Management tab and click Trainee Sign-in. There you can allow trainees to sign in with Google or Microsoft, turn on each provider, and optionally list the email domains you accept. Once it's on and we've finished setting that provider up for your practice, your trainees see a Sign in with Google or Sign in with Microsoft button on the sign-in page. A trainee can sign in this way only if they've already been invited to your workspace — it never creates new accounts, and it applies to trainee accounts only, so owners, admins and hunters keep using their email and password. Signing in with Google or Microsoft replaces the password, not the security code: everyone who opens the TrainBase website enters their six-digit code, however they signed in. In the browser extension, Google or Microsoft on its own is enough. The same page lets you issue, turn off, or replace reusable access keys for the browser extension.

Yes! In the Admin panel under "Share the Blueprint", click "Invite a Colleague". You earn 1 free month for each colleague who signs up, and they receive 50% off for 3 months. A tracking table shows all your referrals with their status, and you can resend or delete invitations at any time.

The Content Gaps tab in the Admin panel lists questions your trainees asked the Training Mentor that your video library couldn't answer — either no videos matched, or the matching videos didn't contain the answer. Each entry shows how many times the question was asked and when. Recording a video on the topic closes the gap: mark it Resolved once the content exists, or Dismiss it if it's not needed. Workspace owners and admins also receive a periodic email digest of new gaps. Questions are automatically privacy-scrubbed before they're stored.

In the Admin panel, click "Upload SOP Document" and select a PDF or text file. TrainBase will analyze the document and generate a list of suggested bounties, each with a title, visual goal, director's note, script prompt, and difficulty level. Review the suggestions, make any edits, then click "Publish to Board" to post them to the Bounty Board.

In the Admin panel's Team section, click the role badge next to any member to change their role. Available roles are Trainee, Hunter, Admin, and Owner. You can only assign roles below your own level, and only the workspace Owner can grant the Admin role — for example, an Admin can promote someone to Hunter, but only the Owner can make someone an Admin.

Onboarding playlists are curated video collections that automatically load for new team members the first time they visit the Training Room. Admins can configure which playlists appear for their workspace, so every new hire starts with the right training from day one.

The Compliance Command Center (available to workspace founders) contains all compliance documentation, including the Security Risk Assessment, incident response plan, breach notification templates, and BAA tracker. Access it from the Admin panel under the Compliance tab.

Yes. TrainBase provides a Business Associate Agreement (BAA) available at /legal/baa. The BAA must be accepted during registration and covers the handling of Protected Health Information (PHI) within the platform. If you need a custom BAA or have questions about specific terms, contact us directly.

Look for the red bug icon button near the bottom-right corner of any page (on pages with a chat panel, it floats just above the chat input). Click it to open the report form. Describe what went wrong, optionally select a category and impact level, then click Submit Report. Your report is sent directly to the TrainBase team.

Click the lightbulb icon button near the bottom-right corner of any page (on pages with a chat panel, it floats just above the chat input). Describe the feature you'd like to see and how it would help your workflow, then click Submit Request. We review every request and use them to guide our development priorities.