FAQ
Every 'yeah, but …' - answered.
In the order buyers actually ask them. If yours isn't here, mail hello@jottings.ai.
About the product.
- A single redacted observation of real work - which app was active, for how long, in what sequence, attached to a workflow fingerprint. Thousands of jottings make a scan. A scan makes a report.
- Those measure individuals for performance review. We measure workflows for automation targeting, and every recommendation carries evidence back to observed usage. Our unit is the process, not the person - and the contract forbids using our data the other way.
- Process mining reads your logs. We read the desk. Logs only show workflows your systems were designed to record. Most of the dollar leak lives in the Excel-Outlook-Salesforce glue between systems - which only desktop observation can see.
- No. They own execution. We tell you which workflow to automate, with which stack, with what ROI. Our reports frequently recommend them by name - and on Firm and above, with your affiliate link.
- A versioned pattern: a workflow fingerprint, recommended tool stacks, implementation hours, HITL thresholds, and outcome-calibrated ROI bands. Jottings matches every scan against the Skill Dojo library.
What exactly is a 'Jotting'?
How is this different from RescueTime, Hubstaff, or time-tracking software?
How is this different from process mining (Celonis, UiPath)?
Do you replace n8n, Lindy, Relevance, Make, or Zapier?
What's a Skill Card?
About the scan.
- Two to four weeks, always. Shorter than that and the fingerprints are unstable; longer and you burn consent fatigue without new signal.
- A minimum of 15 for most engagements. Below that, workflow patterns are too sparse to cluster. Company-tier deployments inside a single business unit can be smaller with our calibration.
- All three. The agent is desktop-first, so geography is irrelevant as long as the consent flow, DPA, and state-appropriate monitoring notice are in order for each jurisdiction.
- Slack and Gmail via OAuth in v1. Jira, Salesforce, HubSpot, and Zendesk in v1.5. Every connector is opt-in and redacted before it hits the synthesis layer.
- Yes. Re-scans are often cheaper because we anchor against your previous work graph and only re-compute deltas. Popular post-automation.
How long does a scan take?
How many employees do you need to observe?
Can we scan remote-only teams? Hybrid? Offshore?
What signal sources beyond the desktop?
Can we re-scan a quarter later?
About privacy and consent.
- No. Bounded, consented, workflow-level, auto-uninstalling. A one-click pause button. An opt-out with no consequence in the DPA. If any of those four go missing, it becomes surveillance - and we refuse to ship that product.
- They never leave the device. OCR runs locally, the extracted text is redacted locally, and the pixel bytes are discarded before network I/O.
- Yes, without consequence. It is a condition of the tri-party DPA the client signed with us.
- No. We measure workflows. Individual identifiers are hashed and unrecoverable by the time a report is rendered.
- NY, CT, DE, and growing. We auto-generate the appropriate notice per engagement based on the employee's state of residence.
- Not in beta. Those verticals come online after SOC 2 Type II and BAAs are in place. We will not cut corners on that timeline.
Is this surveillance software?
What happens to raw screenshots?
Can employees opt out?
Do you track individuals?
Which states have special monitoring notices?
Will this work in healthcare, law, or regulated finance?
About the report.
- Jottings drafts every section from your scan's data. The consultant or internal lead edits, regenerates, and signs it. The consultant is the attesting author on the cover.
- Yes. A split-pane editor with AI-drafted sections on the left, traceable evidence on the right, and inline regenerate. Drag to reorder opportunities. Swap tools per Skill Card.
- On Firm and above - full palette, logo, methodology page, cover. Solo gets logo and color only.
- Each opportunity shows a p25/p50/p75 band, sourced from the observed frequency in your scan × loaded labor cost × a Skill-Card-calibrated automation ratio. Every figure is traceable to its underlying evidence cluster.
- PDF at GA. Keynote and PowerPoint export in v1.5, behind a single click.
Who writes the report?
Can I edit it?
Can I white-label it?
How accurate are the dollar impact numbers?
Can I export to Keynote or PowerPoint?
About pricing and billing.
- A 60-second demo scan runs on your own machine with synthetic data. No credit card. Real scans require an engagement and a signed DPA.
- One complete 2–4 week engagement. Unused credits roll within the annual term.
- Yes - ~2 months free on Firm and Agency paid annually.
- Buy per-scan credits ($500 each) or upgrade. No auto-charges, no 'contact sales' wall.
Can I try it before paying?
What's a scan credit?
Annual discounts?
What happens if I exceed my tier's engagements?
About data and security.
- Supabase, US regions, with engagement-scoped encryption keys. Raw redacted observations land there; the keys are rotated at engagement close.
- 90 days post engagement close, then purged. The deletion generates a hashed attestation log the client keeps.
- Type I in progress, targeted at Month 6. Type II follows.
- Not in beta. Healthcare engagements wait for Type II + BAA.
- Yes. Certificates available on request to company counsel.
Where is data stored?
How long do you retain observations?
Are you SOC 2?
Is there a BAA?
Cyber liability and E&O insurance?
Your first scan tells you more than your last six months of workshops.
Code-signed · auto-uninstalls on end date · opt-out any time.