One solution for all your local IT operations
Healstead watches every endpoint, fixes the safe problems by itself and hands the rest to your team with the evidence ready.
- Every computer, one consoleLive health for Windows and Linux laptops, desktops and servers across all your sites.
- Fixes that run themselves, safelyOnly the fixes your security owner approved run automatically, and each one is checked again afterwards.
- Service desk built inTickets, SLAs, approvals and notifications, with the evidence already attached.
- AI that shows what it checkedDiagnosis cites what it saw. It runs on your own servers unless your administrator turns on an outside AI provider. The AI only suggests; every fix is decided by your signed policy.
- Onboard in minutesDiscover the computers on your network, tick the ones you want and import them together.
Every action approved and recorded · Healstead by DevHeal Labs

Welcome back
Sign in with your DevHeal Labs account, Google or your organisation’s sign-in.
Use a short-lived access token from your company sign-in service. It is kept in this tab only.
IT Operations Overview
Live health and autonomous activity across your managed environment.
Tickets
Incidents and requests, with their status, owner and due time.
New Healstead ticket
Incidents & requests
| Ticket | Title | Computer | State | Priority | Resolution due | Assigned | |
|---|---|---|---|---|---|---|---|
| Connect to load native tickets. | |||||||
Ticket details
Issue and current status
Investigation · analyst review
Unconfirmed analysis: it never runs anything. Check how fresh the evidence is before acting on it.
Conversation with the employee
Activity timeline
Governed operations
Attachments
Related tickets
Employee notifications
Durable in-app ticket updates. External email and webhook delivery can be configured separately.
Endpoints
Laptops, desktops and servers with live health, plus phones, tablets and Macs read from Intune. Select a device for details.
Incidents
Every reported problem and what Healstead found. Causes stay unconfirmed until a fix is verified.
Recent incidents
| Ticket | Problem | Family | Site | Status | Reported |
|---|---|---|---|---|---|
| Connect to view incidents. | |||||
Servers
File and print servers, domain controllers and branch servers: health, inventory and safe fixes.
Autonomous Operations
Every automatic fix from detection to independent verification.
Remediation Library
Versioned fix skills, their risk, autonomy and track record.
Policies & Autonomy
What may run automatically, what needs approval, and where.
Whatever the signed policy says, Healstead never does these: Delete user documents, run AI-written commands, change security or identity: never.
Enrollment & Deployment
Add devices from the console: push on the company network or a one-line command anywhere.
Integrations
Connected systems and their real connection state.
Audit
Read-only record of every decision and action, protected by a hash chain.
Reports
Download incident and device data for one site.
Download a report
The report contains only records you are allowed to see. It is usually ready within a minute.
Governance controls
Independent policy, durable audit, reviewed actions
Audit-chain integrity
Checks the tenant's cryptographic hash chain. A pass is not a substitute for independent immutable off-site storage.
Knowledge
Step-by-step guides for employees and the AI. A second person approves each one before anyone sees it.
Guides
| Title | Family | Platform | Status | Owner / approver | Action |
|---|---|---|---|---|---|
| Connect to load knowledge. | |||||
Submit a guide for review
Write steps an employee can follow safely. Guides that ask for deletion, restarts, installs or commands are rejected automatically. A different reviewer must approve before anyone sees it.
Platform
Every company and account on this Healstead service. Platform owner only; customer tickets and devices stay with each company.
Settings
People, safety switches and company policy. Every change needs MFA, a second administrator where it matters, and is recorded.
Guardrails
Read live from your company's signed fix policy and settings.
Emergency stop
Freezes diagnostic request signing, external ITSM writes and ticket closure. Any scoped admin can engage it; release needs an MFA administrator.
Problem clusters
Bursts of the same incident family at one site, handled as one parent problem. Grouping is not proof of a shared cause.
| Family | Cases | Devices | Max priority | Last seen |
|---|---|---|---|---|
| Choose a site to look for clusters. | ||||
Policy versions
One administrator proposes; a different MFA administrator activates. Automatic closure, device-changing operations and remote support cannot be enabled here.
Settings in force and where they come from
| Setting | Value | Source | Allowed |
|---|---|---|---|
| Connect to load settings. | |||
| Version | Status | Proposed by | Approved by | Settings | Action |
|---|---|---|---|---|---|
| Connect to load policy history. | |||||
Model & knowledge registry
Register AI models, learned policies and knowledge bundles with their checksum, licence and test result. A different administrator approves; retired items stop being trusted.
| Kind | Name | Checksum | Status | Action |
|---|---|---|---|---|
| Connect to load registered artifacts. | ||||
Local AI budget & kill switch
Token budgets cap local-model use per tenant per day and per case. The kill switch sends every investigation to deterministic rules and escalation across all workers.
Control evidence
Live checks mapped to NIST AI RMF, NIST SSDF, OWASP LLM, ISO 27001 and SOC 2. Anything not observable here is shown as not evidenced. This is not a certification.
| Control | Status | Evidence | Frameworks |
|---|---|---|---|
| Run the checks to see results. | |||
Platform health
Healstead checks its own database, queues, cache and storage every five minutes and repairs internal state only. Autonomy per issue is earned and can be lowered here.
| Check | Status | Detail | Autonomy level |
|---|---|---|---|
| Connect to load platform health. | |||
Reviewed learning
Outcomes confirmed by MFA reviewers rank approved guides. Trained policy candidates appear in the registry and need approval before use.
Declare a dependency
Tell Healstead what depends on what, e.g. application "payroll" depends on database "pg-hr".
Import incidents
Paste up to 1,000 rows as CSV with a header: source_id,category,symptom,priority,device_id. Rows are validated individually; duplicates are recognised by source ID.
Selected record
This view contains redacted, read-only case or device details.
Raw record
Select a device or incident to inspect the relevant evidence.