enterprise deployment

Deploy workflow intelligence without a surveillance cloud.

This is the playbook from our standard design partner agreement, the one companies actually sign. Deploy one department on up to 10 machines, capture real work for 30-60 days, and leave with SOPs, an automation map, and a contractual expansion decision. The workflow map is an output of capture, not homework we hand you up front.

how a pilot actually runs

One department. Up to 10 machines. 30-60 days of real work.

Nobody is asked to pick their "one workflow" in a kickoff meeting. People do their jobs; the repeated workflows, exceptions, and escalation points surface from what was actually captured.

design partner pilot · phase 1
01
Kickoff and scope
days 0-2

You name a business sponsor and a deployment owner. Together we scope which apps, sites, and data categories are in or out, and you handle employee notice. A shared support channel goes live at kickoff.

02
Deploy one department
week 1

Up to 10 machines, through the path you already have: Intune, SCCM, a packaged installer, or manual installs. The bar in the agreement: every machine active and capturing within the first week.

03
Capture real work
30-60 days

No scripted demos. People do their normal job while capture runs locally. Repeated workflows, handoffs, exceptions, and escalation points surface from the data. Weekly working sessions and capture-health reviews throughout.

04
Decide expansion
end of phase 1

We review the deliverables with your operators and managers. You make a clear yes/no on expanding to roughly 30 machines. Nothing auto-renews and there is no surprise per-seat billing; expansion happens on a written order or not at all.

phase 1 deliverables

You leave with artifacts, not a slide deck.

Every deliverable below is named in the agreement. SOPs are reviewed with the operators who do the work, and the automation matrix says who should build each candidate: Screenpipe agents, your engineers, an RPA partner, or nobody.

phase-1-deliverables/
workflow-inventory.mdevery repeated workflow, handoff, and exception found in capture
sop-drafts/reviewed line by line with your operators, not generated and dumped
escalation-map.mdwhere humans must stay in the loop, and why
automation-matrix.csveach step mapped to agent, deterministic code, RPA, API, or human
capture-health-reports/device activity and data quality through the pilot
expansion-recommendation.mdwhat to roll out to the next ~20 machines, and what to skip

success criteria

Success is defined in the contract, not in a recap email.

Phase 1 is successful when the boxes below are checked. If they are not, you do not expand, and nothing renews on its own.

All pilot machines active and capturing within the first week

The main repeated workflows identified and summarized

First SOP drafts reviewed together with your operators

A concrete list of automation candidates and SOP gaps

A clear yes/no decision on expanding past the pilot

A named execution layer for every automation candidate

after the map

Screenpipe is the source of truth. The automation goes to the right execution layer.

Captured activity, workflow maps, SOPs, and escalation rules describe how the work happens. Phase 1 produces the evidence to decide who should build each automation, instead of forcing everything through one tool.

Screenpipe agents

Workflows that run directly on captured context: summaries, CRM updates, ticket enrichment, SOP-checked steps.

Your engineers

Work that needs changes to your WMS, routing tools, or internal APIs. The SOP and matrix become their spec.

RPA / computer-use partners

Reliable execution inside third-party portals with no API. The captured traces become their test cases.

Humans, on purpose

Judgment calls and exceptions stay with people, documented in the escalation map instead of automated badly.

deployment modes

Local-first does not mean one data path.

Screenpipe can run as a local-only personal assistant, a scoped team deployment, or an embedded capture engine. The important question for buyers is not a slogan; it is which data flow they approve.

Local-only

What stays local
Screen capture, accessibility text, OCR output, audio files, transcripts, and the local database.
What may leave the device
Nothing is required to leave the device for core capture and search.
Buyer decision
Best for self-serve use, regulated pilots, and proving value before any cloud path is enabled.

Local + optional cloud AI

What stays local
The raw capture store remains on the endpoint unless the user or organization enables export or sync.
What may leave the device
Selected prompts, summaries, or context snippets may be sent to the chosen AI provider or confidential route.
Buyer decision
Buyer chooses model, provider, retention posture, redaction, and whether local models are required.

Team / enterprise

What stays local
Endpoint capture and local history can stay on managed devices under admin policy.
What may leave the device
Team reports, sync, admin workflows, exports, connectors, and agent outputs depend on deployment scope.
Buyer decision
Buyer defines consent, retention, employee controls, report contents, and admin visibility.

SDK / OEM

What stays local
The embedding app defines the storage path, model path, and user-facing privacy controls.
What may leave the device
Data movement depends on the partner architecture and the contractually agreed processing path.
Buyer decision
Partner owns data-flow design, disclosures, user consent, and downstream model/provider choices.

point of view

Capture should produce decisions, not surveillance dashboards.

Screenpipe's enterprise lane is workflow intelligence for AI adoption: prove which work repeats, what can be automated, what an agent should attempt, and which data paths the buyer approves.

One department beats a fleet rollout

Start with up to 10 machines in one team and capture real work. The workflow map is an output of capture, not homework the buyer does up front.

The useful data lives between systems

ERP, CRM, and ticketing logs miss the spreadsheet, tab, message, meeting, and judgment step. That is where the automation target usually hides.

Agents need traces, not vibes

A usable computer-use agent spec needs real inputs, expected outcomes, edge cases, failure modes, and a way to grade the result.

Privacy is part of the deliverable

A workflow report should say what was captured, excluded, redacted, retained, exported, and shared before the team expands deployment.

before kickoff

What we will ask you to settle in the first call.

These map one-to-one to the partner obligations in the agreement. Having answers makes kickoff a day, not a month.

Who is the business sponsor, and who owns deployment?

Which department and which 10 machines go first?

Which apps, sites, and data categories are out of scope?

Local-only capture, or is cloud AI approved, and which provider?

Who handles employee notice and consent?

What can admins see, and what is excluded or scrubbed at the source?

Who approves expansion past the pilot?

Model training on your data: opted out, or explicitly scoped in writing?

the boring terms, up front

You own the captured data, the workflows, and the SOPs.

Model training only under a signed scope table, with opt-out.

Never used to train third-party foundation models without written approval.

Your name is never used publicly without written sign-off.

Nothing auto-renews. Expansion requires a written order.

Weekly working sessions and a shared channel through the pilot.

Scope, timeline, and commercials are settled on the first call, against this playbook.

Plan a pilot