Lapu AI for Teams
Automate the desktop work your team still does by hand.
Lapu AI is a desktop agent that executes multi-step work inside the apps your company already runs — including the legacy ones without APIs. Permission-gated, fully logged, deployed on your machines.
No sales deck. We map 2–3 of your workflows live and tell you honestly whether Lapu fits.
macOS · Windows · VMs
Runs on the machines you already manage — physical or virtual
No API needed
Works with legacy apps through the UI
Per-seat or per-bot
Billing that matches how you run automation
The problem
Every team has work that software was supposed to eliminate.
It didn't disappear — it moved between systems. Someone still re-types, renames, reconciles and reports by hand.
Re-typing between systems
Data moves from email to Excel to your ERP through someone's keyboard.
Reports assembled by hand
The same monthly deck, built from the same six exports, every month.
Legacy apps nothing integrates with
Your core Windows software has no API — so no integration platform can touch it.
Files nobody wants to own
Renaming, sorting, merging, archiving. Necessary, endless, invisible.
This is the work Lapu takes over.
For the organization
What changes when the work runs itself
Throughput without headcount
Recurring desktop work becomes scheduled workflows. The team reviews and approves instead of executing keystroke by keystroke.
No integration project
Lapu works through the same interfaces your people use. No API, no middleware, no six-month rollout. If an employee can do it on screen, Lapu can automate it.
Control you can defend
Sensitive actions require explicit approval. Files stay on the machine. Every step is logged with timestamps and parameters — an audit trail you can show your IT and compliance teams.
Use cases
Where teams put Lapu to work first
Operations
Order intake to ERP, status reports, cross-app data moves.
See workflowsFinance
Invoice extraction to Excel, reconciliation prep, monthly close packages.
See workflowsBack-office & admin
Document processing, folder hygiene, archive migrations.
See workflowsCustomer service
Case data lookups across tools, response drafting from internal files.
See workflowsRollout
From intro call to running workflows in under a month.
Day 0
Intro call — 30 min
We map 2–3 of your recurring workflows and tell you which are automatable. If Lapu isn't the right fit, we'll say so.
Weeks 1–2
Pilot
A fixed-scope pilot: 2–5 users, your real workflows, our help setting them up. You measure the hours. Scope and pricing are agreed on the intro call.
Week 3
Rollout
Standard macOS and Windows installers, compatible with your device management. Runs on physical machines and VMs alike. No server-side infrastructure to deploy.
Week 4
Review & scale
We review pilot results together — hours saved, error rates, and the next workflows worth automating.
Step 1 costs 30 minutes. Everything after it is your call.
Security & compliance
Built to pass your IT review.
An agent with access to files and desktop apps has to clear a higher bar than a chatbot. That's how Lapu is architected.
Permissioned execution
Every sensitive action requires explicit user approval before proceeding. File writes, shell commands, and desktop automation are gated by default. The agent never acts beyond its granted scope.
Local-first architecture
Files and data remain on the user's machine. There is no Lapu AI cloud storage, no remote workspace, and no ambient data collection. When AI reasoning is needed, only the relevant context is sent to the model endpoint managed by Lapu AI.
Full audit trail
Every agent action is logged and visible in real time. Tool invocations, file operations, and model calls are recorded with timestamps, parameters, and outcomes for complete operational transparency.
Context isolation
The agent runtime separates renderer, backend, and system-level processes. Each component operates within defined boundaries to limit the scope of any single action.
What leaves the machine
- Your files
- Stay on the machine. No Lapu cloud storage.
- Task context
- Only the relevant excerpt is sent to the model endpoint, per task.
- Action log
- Stored locally, exportable for audit.
Coverage
Modern stack, legacy stack — same agent.
Automates work across the tools your team already uses — right from your desktop.
- Google Workspace
- Microsoft 365
- Salesforce
- HubSpot
- Notion
- Jira
- GitHub
- Figma
- Zoom
- Stripe
- Asana
- Airtable
- Linear
And the apps with no logo on this strip.
The strip above is the easy part. Lapu also drives the software integration platforms can't reach — the 15-year-old Windows app your business actually runs on. If it has a UI, it can be automated.
New product, honest terms.
Lapu is new. Instead of borrowed logos, we offer proof on your own workflows: a two-week pilot on your real processes, with hours measured against your baseline. If the numbers don't work, you'll know in 14 days.
The business case
What is manual work costing you today?
Your team size, your hours, your rates — the calculator only multiplies numbers you give it.
What manual desktop work costs you per year
$92,000
Your numbers × 46 working weeks. Nothing assumed, nothing inflated.
Qualification
A good fit — and an honest miss.
Lapu is a strong fit if:
- Your team runs recurring, rule-describable desktop work
- Key software has no API — or an expensive one
- Work spans several apps and someone is the glue between them
- You need control and logs, not a black box
Lapu is probably not for you if:
- Your entire stack is cloud tools with solid APIs — an integration platform may be cheaper
- The work requires judgment calls a human must own end to end
- You need a fully on-prem language model today
Sounds like you? . Not sure? Book it anyway — qualifying honestly takes us 10 minutes.
For the decision maker
The questions procurement will ask
How does deployment work?
Lapu AI ships as standard macOS and Windows installers and runs on each user's machine — physical or virtual. VMs are a supported deployment target, so bots can run on dedicated virtual machines. There is no server-side component to stand up; rollout works with your existing device-management tooling, and most teams are running within days, not quarters.
How is Lapu licensed for teams?
The Teams plan includes a shared usage pool, unlimited seats, a team management console, and centralized billing. Enterprise adds SSO/SAML, a custom SLA, and custom deployment. If you prefer the model you know from RPA platforms, per-bot billing is available too — you license each bot instead of per-user seats. Both are priced to your scope — that conversation starts on the intro call.
What data leaves our machines?
Files and workspace data stay local — there is no Lapu cloud storage. When a step needs AI reasoning, only the relevant context for that step is sent to the model endpoint. Every action is written to a local, exportable audit log. The full architecture is documented on our security page.
Can we control what the agent is allowed to do?
Yes. Sensitive actions — file writes, shell commands, desktop automation — are gated behind explicit approval prompts by default, and every action lands in the audit trail. The agent never acts beyond the scope it has been granted.
What does the pilot involve and what does it cost?
A two-week, fixed-scope engagement: 2–5 users, 2–3 of your workflows, hands-on setup help from us, and hours measured against your baseline. Pilot pricing depends on scope and is agreed on the intro call — no long-term commitment required.
What support do we get?
Teams plans include priority support; Enterprise adds a dedicated account manager and a custom SLA. During the pilot you work directly with us on workflow setup.
How do you handle our security and data-processing requirements?
Bring them to the intro call. We walk through the data-flow architecture, the permission model, and the audit trail with your IT or compliance stakeholders, and handle data-processing terms as part of procurement.
Question we didn't answer? .
Bring us one workflow. We'll show it running.
A 30-minute call: you describe 2–3 processes, we tell you what's automatable, you leave with a pilot plan — or an honest "not yet".
30 minutes · no obligation · we'll tell you if it isn't a fit
- 01Pick a slot that suits you
- 02No preparation needed
- 03Summary & pilot proposal within 24h
Prefer to try it yourself first? Download the desktop app →