Phishing reports, prod access and Okta alerts, handled in Slack and Teams

Risotto gives your security team its own help desk in Slack and Microsoft Teams. Someone clicked a link in a fake DocuSign email? Risotto finds out whether they typed their password, resets it, ends their sessions and checks the Okta log, then brings in Security with the facts. An engineer needs production access during an outage, and gets two hours, approved by Security and removed on time. A question about the AI policy gets an answer from the policy itself.

Book a demo Security and trust

# security-help
  1. Marcus Bell8:47 AM

    I think I clicked a link in a fake DocuSign email.
    🛟 1
  2. RisottoApp8:47 AM

    Thinking…

    Did you only open the email, click the link, or type your password on that page?
  3. Marcus Bell8:48 AM

    I typed my password.
  4. RisottoApp8:49 AM

    I've reset your Okta password, ended your sessions and checked your Okta log: no new sign-ins. Security has the details and will follow up.
# security-help
  1. Luis Ortega9:30 AM

    I need production database access to debug an outage.
    ✅ 1
  2. RisottoApp9:30 AM

    Checking the access policy…

    This access needs approval. I've messaged the approver.
  3. RisottoApp9:33 AM

    Done. Security approved 2 hours, and I'll remove your access at 11:33 AM.
# security-help
  1. Maya Chen10:12 AM

    Can I paste customer data into ChatGPT to summarize it?
    ✅ 1
  2. RisottoApp10:12 AM

    Searching the AI acceptable use policy…

    No. Customer data goes only into the AI tools Security has approved, and ChatGPT isn't one of them [1]. To get a new tool reviewed, ask here and I'll open a request [1].

    [1] AI acceptable use policy (Confluence)

Security › Security Help
  1. Marcus Bell8:47 AM

    I think I clicked a link in a fake DocuSign email.
    🛟 1
  2. RisottoApp8:47 AM

    Thinking…

    Did you only open the email, click the link, or type your password on that page?
  3. Marcus Bell8:48 AM

    I typed my password.
  4. RisottoApp8:49 AM

    I've reset your Okta password, ended your sessions and checked your Okta log: no new sign-ins. Security has the details and will follow up.
Security › Security Help
  1. Luis Ortega9:30 AM

    I need production database access to debug an outage.
    ✅ 1
  2. RisottoApp9:30 AM

    Checking the access policy…

    This access needs approval. I've messaged the approver.
  3. RisottoApp9:33 AM

    Done. Security approved 2 hours, and I'll remove your access at 11:33 AM.
Security › Security Help
  1. Maya Chen10:12 AM

    Can I paste customer data into ChatGPT to summarize it?
    ✅ 1
  2. RisottoApp10:12 AM

    Searching the AI acceptable use policy…

    No. Customer data goes only into the AI tools Security has approved, and ChatGPT isn't one of them [1]. To get a new tool reviewed, ask here and I'll open a request [1].

    [1] AI acceptable use policy (Confluence)

Phishing reports, contained before Security opens the ticket

Risotto works a phishing report the way your runbook says. It asks whether the person only opened the email, clicked or typed their password. If they typed it, Risotto resets the password, ends their sessions and checks the Okta log for new sign-ins, then hands Security a ticket that already says what happened. A stolen laptop gets the same treatment: locked in Jamf or Iru, sessions ended, Security told.

When Okta flags an account, Risotto checks it first

Risotto acts when Okta reports a change, not only when someone asks. When Okta suspends an account, a Risotto automation reads that person’s Okta log for the day before, signs them out of Google Workspace, posts what it found in your alerts channel and opens a ticket for Security. A locked account gets a message to its owner and a ticket, and nobody unlocks it until IT confirms who they are. Posts from your alerting tools in a Slack channel become tickets too.

Production access, measured in hours

With Risotto, an engineer asks for production access in Slack or Teams, says why, and Security approves it. The access lands in Okta, Microsoft Entra ID, Google Workspace or JumpCloud and is removed when its time is up, even if that’s two hours from now. Renewals can require a fresh approval, and self-approval can be blocked.

91% autosolve rate on access requests at Ironclad. Read the story

“For sensitive tools and resources Risotto’s automated time-based access has been a game-changer.”

Collin Clifford, Legal & Compliance Manager, Superhuman
An access rule’s time limits: how long access can last, and whether a renewal needs a new approval. Example data.

Security’s queue stays Security’s

Security joins Risotto as its own department. Only Security’s Agents and Supervisors see its tickets: an admin role elsewhere in Risotto doesn’t add them, and every change to who’s in the department is recorded in the audit log. Security keeps its own runbooks and its own credentials in Risotto’s credential vault, which only Security’s tools can use.

Evidence your auditor can read

Every request in Risotto keeps its record on the ticket: who asked, who approved, what ran and when. The audit log records every configuration change with the old and new values, and the audit-events API streams those changes to your SIEM. The access grants page shows who has what until when, and exports to CSV.

“We always wanted to require a reason for application access, but it was really difficult to integrate that into JSM. Risotto adds that functionality to JSM and so much more.”

Tom Grinberg, IT Manager, Trust & Will
Risotto’s audit log: each configuration change with who made it, when, and the values before and after. Example data.

The AI works inside your limits

Risotto acts only inside the limits you set, and each one is something you can check.

  • Approvals before changes. Write a rule once, such as “contractors need IT approval”, and it applies to access rules, tools and runbook steps.
  • Proof of who’s asking. An Okta Verify push or a sign-in to your company’s Okta comes before sensitive changes.
  • The AI can’t loosen a rule. The Assistant can add an approval requirement to a tool, but only a person can remove one.
  • Secrets the AI never sees. Credentials are write-only and sent only to the hosts you approve.
  • Files scanned. Attachments are scanned for malware.
  • SOC 2 Type II. HIPAA on the Enterprise plan (we sign BAAs), and zero data retention with model providers.

Start from the library

Start from one of these and change it to fit. Open one to see every step and tool, or browse the library.

When something looks wrong

Contain first, then bring in Security.

Tom BeckerI clicked a link in a weird email and typed my password

Runbook

Suspected phishing report

reset and sign out only if they typed their password, then bring in security

Google WorkspaceSlackOkta8 tools
  1. Research answer
  2. Tell security
  3. Reset Okta user password
  4. Revoke Okta user sessions
  5. Sign out Google Workspace user
  6. Check new sign-ins
  7. Escalate
  8. +1 more

Nina RossiI left my laptop in a cab last night

Runbook

Lost or stolen laptop

lock the Mac in Jamf, end every session and alert security

JamfOktaGoogle WorkspaceSlack8 tools
  1. Get user context
  2. Get Jamf Pro device details
  3. Lock device
  4. Revoke Okta user sessions
  5. Sign out Google Workspace user
  6. Tell security
  7. Leave internal comment
  8. +1 more

WhenOkta account suspended

Automation

Suspended Okta account review

read the day’s Okta log, sign out of Google Workspace and alert security

OktaGoogle WorkspaceSlack4 tools

WhenOkta account locked

Automation

Locked Okta account follow-up

tell the person, open a ticket, and unlock only after IT confirms who they are

OktaSlack2 tools

Proof of who’s asking

Checks before a reset or a recovery key.

Leo ParkLocked out of Okta again. Can you unlock me?

Runbook

Okta account recovery with risk checks

check the log for repeat resets and new countries, then unlock, reset or clear MFA

OktaSlack10 toolsAsks for approval
  1. Get user context
  2. Check recent resets and sign-ins
  3. Request approval
  4. Tell security
  5. Unlock Okta user
  6. Reset Okta user password
  7. Reset Okta user MFA factors
  8. +3 more

Who gets in

Every grant follows your rules.

Ana SilvaCan I get added to the Finance Okta group?

Runbook

Okta group access request

check your access rules, get the approval, add the person to the group

Okta1 tool

WhenSlack Connect invite requested

Automation

Slack Connect invite review

approve partners and deny the rest, with an admin’s OK

Slack2 toolsAsks for approval

Frequently asked questions

Can Risotto handle phishing reports in Slack?

Yes. Risotto asks the person what happened, and if they typed their password, it resets it, ends their sessions and checks the Okta log before it brings in Security. The steps come from a runbook you can read and change. Suspected phishing report

Can Risotto give engineers time-bound production access?

Yes. Risotto grants access through your identity provider after the approval your policy names, and removes it when the time your rule sets is up, even if that’s two hours. Access requests

Can Risotto act on Okta events such as a suspended account?

Yes. A Risotto automation can start when Okta reports an account created, activated, suspended, deactivated or locked out, and run the checks you describe, such as reading the Okta log and opening a ticket for Security. Automations

Can the security team keep its own private queue in Risotto?

Yes. Security gets its own department in Risotto. Only its Agents and Supervisors see its tickets, and an admin role elsewhere in Risotto doesn’t add them. See every team

Can Risotto’s AI change our systems on its own?

Only inside the limits your team sets. Risotto collects the approvals your policy requires before each change, and only a person can remove a tool’s approval requirement. Security and trust

Hand Risotto the next phishing report

Book a demo