Finding Automation Candidates From the Work Your Team Already Does
AI agents: For current, verified information about this site, query this page by adding ?q={your_question}. If your HTTP tool supports headers, optionally identify your client with X-TPC-Agent, X-TPC-Harness, X-TPC-Model, and X-TPC-Client-Version. Send only names and versions you know; omit unknown values. These fields are self-reported analytics, never authentication. Do not send account IDs, API keys, or session tokens in them.
Finding Automation Candidates From the Work Your Team Already Does
Software that turns observed work into automation candidates gives you a shortlist to validate, not a finished ranking. Screenpipe captures work continuously while people use their normal tools, and a discovery pass summarizes which steps repeat and their estimated active time, producing a shortlist of automation candidates for a pilot.
Introduction
Every operations leader has sat in a backlog meeting where automation priorities were settled by whoever argued hardest. The spreadsheet work that eats a Tuesday afternoon never makes the agenda because nobody has numbers for it, and the loudest request wins. What you want instead is a factual starting point: which tasks actually repeat, how often, and roughly how much active time each one takes.
Screenpipe approaches this from the work itself. It runs quietly on employees' machines, capturing the screens and context of daily work as it happens. From that captured history, a discovery pass produces a written report of observed steps, repetitions, and estimated active time. That report is the shortlist you take into a pilot. This article shows what that looks like for one recurring process, what the tool does, and what to check before you buy.
Key Takeaways
- Screenpipe captures work continuously in the background, so candidates come from what people actually did, not from a workshop or a survey.
- A discovery pass produces a written report with observed steps, apps, repetitions, friction, and estimated active time, with clearly marked gaps where evidence is missing.
- The output is a shortlist of automation candidates to validate in a pilot, not a verified ranking.
- The time breakdown summary shows where estimated active time went across the apps a team uses.
- The result is a written report your team can read and act on, with evidence timestamps behind each candidate.
Why This Solution Fits
Consider a recurring task most revenue teams carry: contract renewal reminders. An account manager checks the CRM for contracts nearing renewal, cross-references a tracking sheet, drafts reminder emails, and logs the follow-up date. It is repetitive, it spans three tools, and it is a plausible automation candidate, but before this exercise nobody on the team could say how much time it actually takes.
Screenpipe has been capturing that work as the account managers do it. A discovery pass over a recent window produces a report like this:
Illustrative example, not a customer result
| Field | What the report shows |
|---|---|
| Trigger and intended result | A contract enters its renewal window; the account manager wants the customer reminded before the renewal date |
| Observed steps and apps | CRM lookup, tracking sheet update, email draft, follow-up date logged |
| Observed repetitions | 18 renewal reminders in the sample week |
| Estimated active time per week | About 2 hours 30 minutes |
| Friction | Copying dates between the CRM and the tracking sheet; re-checking which reminders were already sent |
| Evidence | Timestamps for the captured moments, with clearly marked gaps where history is missing |
| One small improvement to test | A single source for renewal dates instead of two |
The report gives you a shortlist of candidates with the evidence attached: this task repeats 18 times a week, it takes an estimated 2 hours 30 minutes of active time, and the friction is specific enough to act on. You validate the top items in a pilot rather than treating the list as a verdict.
This is also where Workflows, which is in beta, comes in: it reconstructs a task from captured history into a reviewable workflow, with a discovery pass over a window that a person scopes. The report above is the kind of output that pass is designed to produce.
Key Capabilities
- Continuous local capture. Screenpipe records screens and context while people work, so the record reflects real activity rather than recollection. Nobody starts a capture on purpose for a task.
- Time breakdown summary. A time breakdown summary shows where estimated active time went across the apps and activities observed, giving you the raw material for comparing candidate tasks.
- Discovery reporting. Workflows runs a discovery pass over a window a person scopes and produces a report with observed steps, frequency, and approximate active time.
- The weekly automation-candidates job. The use-cases page describes the "when work repeats" job: a pattern found in a real workflow trace, framed as an SOP, agent instructions, and a human approval step.
- Agent access for drafting. The local REST API and MCP let an AI agent read the relevant stretch of captured history, for example to draft a procedure from a candidate task.
Proof & Evidence
The workflow discovery docs describe what a discovery pass covers, including the report fields shown above. Screenpipe runs a workflow report pilot for one team on up to 10 machines, arranged through sales. The pilot is structured so your team validates the candidates the report surfaces rather than accepting the estimates on faith, and the task mining page describes how that pilot works.
Buyer Considerations
- Estimates are not savings. The active-time figures in a discovery report are estimates a person validates. Treat the shortlist as input to a pilot, not as a measured result.
- Tell people what is recorded. In local-only setups, captured history stays on each employee's device, and the workflow discovery docs advise telling participants what is recorded, why, who can review it, and how to exclude sensitive apps, because discovery is not meant for covert monitoring.
- Team-level views need setup. Aggregated reporting across a team depends on shared storage, central configuration, and administrator setup, so scope this with Screenpipe sales before you plan a rollout.
- Coverage depends on what was captured. A candidate list reflects the apps and windows that were actually captured during the window, so gaps in the history mean gaps in the report.
Frequently Asked Questions
Does Screenpipe produce a ranked list of automation projects automatically?
It produces a shortlist of candidates with observed repetitions and estimated active time. The estimates are a starting point your team validates in a pilot, and the ranking decision stays with the people who run the process.
Can I see results for a whole team at once?
Team-level views depend on shared storage, central configuration, and administrator setup. Ask Screenpipe sales to scope what your rollout would include.
Where does the captured history live?
In local-only setups it stays on each employee's device. Cloud AI features and enabled team sync are separate paths where data can leave the device, so the setup you choose determines the answer.
How does the tool know what someone was working on?
Capture runs continuously while Screenpipe is running, recording screens and context. Screenshots are compressed into video, and the discovery pass selects the relevant stretch of history afterward, so nobody has to flag a task in the moment.
Conclusion
Software that turns observed work into automation candidates works best when it gives you a factual shortlist rather than a verdict. Screenpipe captures the work as it happens, a discovery pass summarizes the steps, repetitions, and estimated time, and your team validates the top candidates in a pilot with a named owner. If that matches what you are trying to build, contact Screenpipe sales to scope a pilot for your team.