Documenting operational knowledge and making it usable with AI
Develop IT operations documentation, keep it versioned in Markdown or a wiki and use it with AI.
AI for IT & Operations at FlixConsulting comprises three components: AI-assisted search across your operations documentation, the creation of comprehensive IT operations documentation that lets your teams work on operational topics directly with AI, and storing operational know-how in a versioned Markdown or wiki structure. It is for IT leaders and operations managers, including in regulated environments.
What changes for you
Answers from your own documentation
Your teams ask questions about operations and get answers with a reference to the source document, instead of searching several repositories one by one.
Operational knowledge does not depend on individuals
Knowledge that currently sits with individuals is documented and versioned. Cover, onboarding and handovers to service providers can be planned.
One source for people and AI
Your teams and the AI system use the same Markdown or wiki pages. Every change is traceable, and every answer can be checked.
What we take on
Available individually or as a package, from analysis to handover.
- 01
Inventory of operations documentation
Capture of existing documents, repositories and knowledge holders, for example in Confluence, Jira or NetBox. Assessment of currency, gaps and inconsistencies for each service.
- 02
Structure, templates and standards
Organisation by services, systems and procedures, templates for operations manuals, runbooks and incident analyses, and mandatory fields such as owner, last updated and validity.
- 03
Creating the IT operations documentation
Operations manuals, runbooks, escalation paths and incident analyses, developed from existing material and interviews with your teams, clearly structured for people and AI.
- 04
Versioned operational knowledge in Markdown or a wiki
Stored as Markdown in a repository, for example in GitLab, or in a wiki such as Confluence. Every change is versioned, reviewed before approval and remains traceable.
- 05
AI-assisted document search
Ask the operations documentation questions in your own words and get answers with a reference to the source document. Only what the person asking may open is shown.
- 06
Working on operational topics with AI
Clarify operational questions, find the right runbooks and narrow down incidents using documented procedures and earlier analyses. Your team makes the decision.
- 07
Protection needs, operating location and permissions
Review of the content against your requirements, for example from ISO 27001, BSI or KRITIS regulations. Then definition of operating location, data flows, read access and logging.
- 08
Maintenance process and handover
Updating the documentation as a fixed step in change and incident management, roles for review and approval, and briefing your teams on documentation and AI use.
What you end up with
- IT operations documentation with operations manuals, runbooks, escalation paths and incident analyses
- Versioned Markdown or wiki structure with templates, mandatory fields and named owners
- AI-assisted document search with source references that inherits existing access rights
- Maintenance process in change and incident management that keeps the documentation up to date, with a briefing for your teams
How we work
- 01
Analysis
We capture existing documentation, repositories and knowledge holders, assess currency and gaps, and check the protection needs of the content.
- 02
Target architecture
We define the structure, templates and storage of the documentation, along with versioning, permissions and the operating location of the AI-assisted search.
- 03
Delivery
We create the operations documentation with your teams, transfer it into the Markdown or wiki structure and set up the AI-assisted search.
- 04
Handover to operations
We embed maintenance in change and incident processes, name owners and brief your teams on documentation and AI use.
- Confluence
- Jira
- GitLab CI/CD
- Python
- Kubernetes
- NetBox
In practice
Questions about this service
How does AI-assisted document search work?
Your teams ask questions in their own words, for example about a procedure or an incident. The search finds relevant passages in runbooks, operations manuals and concepts and summarises them with a reference to the source document. This means every answer can be checked. Only documents that the person asking is allowed to open are shown. The quality of the answers depends on the quality of the documentation.
Why is operational knowledge stored in Markdown or in a wiki?
Markdown and wiki pages are clearly structured and can be versioned. Every change is traceable by author and date and is reviewed before approval. Your teams and the AI system read the same pages. Whether it is stored in a repository, for example in GitLab, or in a wiki such as Confluence, we decide with you based on your existing tools and requirements.
Can AI-assisted search run in our own infrastructure?
Yes, that is possible, and with high protection needs it is often a prerequisite. The search then runs in your data centre, for example on Kubernetes, and the documentation does not leave your network. Before the start, we clarify which operating location is permitted based on your requirements, for example from ISO 27001, BSI or KRITIS regulations. External services are only an option if these requirements allow it.
Let us talk about your project.
No obligation, confidential and with a concrete next step.