Skip to content

/articlesIT operations

ServiceNow incidents, CMDB and knowledge through conversation

Botify reads ServiceNow incidents, problems, CMDB links, changes and knowledge articles for your IT team. Opening or updating an incident waits for approval.

Updated
Published
min read
07
On this page
08
On this page08

What can Botify do in ServiceNow?

Botify saves your on-call engineer the first half hour of an incident. That time usually goes on the same lookups: which incidents are open for the group, which server or application is affected, what depends on it, what changed recently and whether a known fix exists. Botify runs those lookups in ServiceNow and posts one short summary with record numbers.

It also keeps the ticket up to date. When the engineer agrees with the diagnosis, Botify proposes the state change and a work note, or proposes a new incident for an alert nobody has logged. Both are writes to your system of record, so both wait for a person to approve.

The ServiceNow integration has 11 actions. Nine only read and run straight away. The two that write always wait for approval.

ActionQuestion it answersApproval
servicenow.search_incidentsWhich incidents are active, resolved or closed for this priority and group?Runs straight away (read)
servicenow.get_incidentWhat exactly does INC0012345 say?Runs straight away (read)
servicenow.search_problemsIs there an open problem record for this pattern?Runs straight away (read)
servicenow.get_problemWhat root cause and workaround does this problem record give?Runs straight away (read)
servicenow.get_ciWhich configuration item is this, by name, sys_id or correlation ID?Runs straight away (read)
servicenow.get_ci_relationshipsWhat sits above and below this item in the CMDB?Runs straight away (read)
servicenow.get_service_cisWhich items does this business service run on?Runs straight away (read)
servicenow.list_changes_near_timeWhich changes were planned to start around 14:05?Runs straight away (read)
servicenow.search_knowledgeIs there a knowledge article for this error?Runs straight away (read)
servicenow.create_incidentOpen an incident with a short description, priority and groupAlways waits for approval (write)
servicenow.update_incident_stateMove an incident to in progress, on hold, resolved or closed, with a work noteAlways waits for approval (write)

Example: triaging a priority 1 incident

A configuration item (CI) is any record in the CMDB, ServiceNow’s map of your servers, applications and services. Here is how Botify triages a P1 for the payments team. Every line in its summary points to a record the engineer can open.

  • Search active P1 incidents for the payments on-call group and read the newest one in full: INC0012345, “Checkout errors since 09:40”.
  • Find the affected item, payment-service, and look one level up and down its CMDB links.
  • Find the business service it runs under, so the summary says “Online checkout is affected”, not only “a server is unhappy”.
  • List changes planned within 60 minutes of 09:40: CHG0031, a database setting approved for 09:20.
  • Search problem records and knowledge articles with the error text for a known fix.
  • Post one answer: likely cause, the record numbers behind it, the affected service and a proposed next step.
  • If the engineer agrees, propose moving INC0012345 to in progress with a work note. Nothing changes until someone approves.

Example: opening an incident from a monitoring alert

Alerts from Dynatrace or Splunk often arrive without a ticket. Botify reads both the monitoring tool and ServiceNow, so it first checks whether an incident already exists for the same item. One outage stays one ticket, not three tickets in three queues.

When Botify proposes a new incident, the approver sees the short description, priority and group before anything is written. The request also carries a unique key that Botify stores in the incident’s correlation_id field. If the connection drops after ServiceNow created the record, Botify finds that incident on retry instead of opening a second one.

Example: “has this happened before?”

Botify searches problem records and knowledge articles with the error text, reads the closest matches and quotes the workaround, with the earlier incident numbers on the same item.

For “what else depends on this database?”, Botify walks the CMDB one step at a time: the item’s parents and children, then the items behind each business service. Every link in its answer is a relationship record the engineer can open.

If your instance has ServiceNow’s MCP Server Console, you can also add its MCP server to Botify like any other MCP server. Until an admin reviews its tools, each call waits for approval.

How does Botify keep ServiceNow data safe?

Every write is a proposal. Reading an incident harms nobody; closing 50 incidents because a model misread a thread does. So the two write actions always stop and wait for a person on your team. The request appears in Botify’s Inbox, and in the chat itself when your team works in Telegram or WhatsApp.

The write actions accept only a fixed set of values. Updating an incident takes one of four states: in_progress, on_hold, resolved or closed. The model cannot invent a state code. Search filters take fixed lists and bounded numbers, and the group field rejects the ^ character, which would otherwise let extra conditions slip into a ServiceNow query.

ServiceNow’s own permissions still apply. Botify connects over OAuth, so every request runs as the connected ServiceNow account and its access rules. Give that account only the roles Botify needs. Your instance may add rules of its own: ServiceNow’s standard lifecycle, for example, asks for a reason when an incident goes on hold.

  • Reading runs straight away; opening or updating an incident always waits for a person.
  • The approver sees the action, a one-line summary and the exact values it will write.
  • Every action, approval and result is kept in Botify’s audit log with the conversation that caused it.
  • Your ServiceNow login details are stored encrypted in your workspace and are never shown to the model.

Where does ServiceNow triage fall short?

Botify is only as good as your CMDB. If relationships are out of date or no business service map exists, Botify cannot say who is affected. ServiceNow’s Common Service Data Model (CSDM) describes how to link technology to the business services that use it. The more of it you fill in, the better Botify’s impact statements.

Knowledge search matches words, not meaning. It looks for your text in article titles and bodies, so it can miss an article that describes the same fault in other words. Ask Botify to try the error code, the product name and the symptom separately.

Changes are matched by planned start time. A change that overran or started early can fall outside the window. Widen the window, up to a full day, when the timeline is unclear, and treat “no changes found” as weak evidence.

How to start with Botify on ServiceNow

Start read-only with one group and a short list of questions. Check that the summaries are right, then turn on the write actions with approval.

  • Create a dedicated ServiceNow account with read access only, and connect it over OAuth.
  • Use only the read actions for the first weeks, and compare Botify’s triage notes with what your engineers concluded.
  • Fix the CMDB gaps Botify exposes. Missing links show up fast once something reads them on every incident.
  • Turn on opening and updating incidents, and decide who approves for each team.
  • Read the audit log each week. Rejected proposals show where Botify’s reasoning or your data is weak.

Common questions

01

Does Botify replace ServiceNow Now Assist?

Not necessarily. Now Assist works inside ServiceNow. Botify helps when the investigation also needs data from Dynatrace, Splunk or Gmail, outside the instance. Many teams use both.

02

Can Botify close incidents on its own?

No. Updating an incident’s state always waits for a person to approve the exact state and work note. Closing a ticket tells the business the service is back, so a person should own it.

03

Which ServiceNow data does Botify need?

For triage: incidents, problems, change requests, configuration items and their relationships, and knowledge articles. Grant read access to those first. Add write access to incidents only when you turn on the approval-gated actions.

04

Will Botify create duplicate incidents if the connection drops?

It should not. Botify stores a unique key in the incident’s correlation_id field and looks for it before trying again, so a retry finds the incident that was already created.

05

Can Botify use ServiceNow’s MCP Server Console?

Yes, if your instance has it. ServiceNow says it needs Zurich or later and its own entitlement. Add its MCP server to Botify by its HTTP address. Its tools then wait for approval on every call until an admin reviews them. Botify’s 11 built-in ServiceNow actions use the standard Table API and do not need it.

Sources

  1. [1]ServiceNow Docs: Life cycle of an Incident
  2. [2]ServiceNow Docs: CSDM implementation stage — Run
  3. [3]ServiceNow Community: MCP Server Console FAQ

Where you’ll find this in Botify

Start with one assistant on WhatsApp or your website.

Book a demo and we’ll show you an assistant answering questions like your customers’. We set up the first one with you and review the results together before you add more.