IT managed service provider ticket archives, 19 MSPs

Summary

Service tickets from 19 US managed service providers supporting small and mid-size business clients, with every customer note, internal note, time entry narrative and resolution. Tickets cover Microsoft 365 and identity issues, endpoints, networking, backups, user setup and removal, and security alerts, which suits IT agents that must diagnose, act and write up the fix. Passwords, keys, MFA codes, license keys, IP and MAC addresses, end-user names and client names are removed, and tickets from healthcare and law firm clients are excluded entirely.

Collected to order. We collect it from US companies for your order. Scope, volume and timing are confirmed in writing before you sign.

Pricing on request

Request scope and pricing

At a glance

Planned scope. Volume, sources and timing are confirmed in writing before you sign.

Records: 6.2Mservice tickets with notes
Size: 15.4 GB
Tokens, estimated: 2.8B
Years: 2016 to 2026
Firms: 19

Owner profile: 19 US IT managed service providers, 6 to 75 staff each, serving clients of 10 to 500 seats
Industry: Professional services
Data type: Support and service
Headcount: 6 to 75 staff per firm
Region: United States
Languages: English
Formats: JSONL
Updated: 2026-10-21

Sources

System	Kind	Records
ConnectWise PSA	Support tickets	4.1M tickets
Autotask PSA	Support tickets	1.6M tickets
HaloPSA	Support tickets	500K tickets

Good for

training IT support agents to diagnose and resolve identity, network and endpoint issues
ticket triage, priority setting and board routing
evaluating resolution write-ups and customer-facing ticket replies
mining recurring root causes across client environments

Fields

Field	Type	Description
ticket_id	string	Stable ticket id
msp_code	string	Code of the contributing firm
client_ref	string	Client code, stable across tickets
board	enum	service_desk, projects, alerts or onboarding
type	string	Type and subtype as configured, normalized
priority	enum	p1, p2, p3 or p4
opened_at	timestamp	Open time in UTC
closed_at	timestamp	Close time in UTC
sla_met	boolean	Whether response and resolution targets were met
notes	array<string>	Customer and internal notes in order, with time entry narratives
resolution	string	Resolution text as closed

Example record

{
  "ticket_id": "MS06-T-1184420",
  "msp_code": "MS-06",
  "client_ref": "[COMPANY_57]",
  "board": "service_desk",
  "type": "M365 / Outlook",
  "priority": "p3",
  "opened_at": "2025-03-04T14:12:09Z",
  "closed_at": "2025-03-04T16:58:41Z",
  "sla_met": true,
  "notes": [
    "Customer: [PERSON_44] says Outlook keeps asking for a password since this morning, webmail works fine",
    "Internal: conditional access policy blocking legacy auth went live last night, this PC runs a 2016 MSI build of Outlook",
    "Internal: cleared cached credentials, installed current M365 Apps, rebuilt the Outlook profile, 1.0 h",
    "Customer: confirmed working, signed in with the MFA prompt"
  ],
  "resolution": "The old Outlook build could not use modern auth after the conditional access change. Upgraded to current M365 Apps and rebuilt the profile. Flagged 6 more MSI installs at this client for the same fix."
}

One record, de-identified. Placeholders such as [PERSON_12] and [COMPANY_4] stay the same across the whole dataset, so threads still read correctly.

Licenses available

Exclusive
One buyer. Nobody else can license it during the exclusive term, usually 2 years.
Non-exclusive
Licensed to more than one buyer.
First look
One buyer gets 2 to 4 weeks to make an offer before anyone else sees the dataset.

What ships with it

Datasheet: what it is, what it is good for, verified size, sources, years, format and limits.
Provenance file: the owner shown by code, staff notice and consent status, exclusions, mapped to the EU AI Act Article 53 training-data summary and California AB 2013.
Manifest with a hash for every file.
QC report and de-identification report.
Your own copy, with unique canary records.

How it is de-identified

Default exclusions removed, files flattened to text, identifiers found in three layers and replaced with consistent placeholders, and the result measured against a 99% bar for each identifier type in each source.

The full method