Every part of service operations, built around the incident.
From the agent that runs first response in Slack and Teams to problems, changes, CMDB, tickets, tasks, drills and regulatory reporting, everything runs on the same record. We build our own AI model and run it alongside frontier models.
SOC 2 Type 1 · ISO/IEC 27001 · Self-Hosted · In-house AI model
Just a work email and company name. No sales calls.
AI
We didn't bolt features onto a general-purpose AI. We're building AI for incident response.
At the core of Incident Lake is Otter, our own model built for incident response. It works alongside frontier models, each doing what it does best, so you get accuracy neither reaches alone and data boundaries neither can keep alone.
Inputs
Slack / Teams
Monitoring
CMDB
Knowledge
Models
Otter
In-house, built for incidents
Routing / boundary settings
Frontier models
Broad knowledge, long context
Outputs
First update
Tasks
Reports
Analytics
01 Otter
A model that speaks your on-call language
It understands what “mitigation,” “affected service” and “alerts settling” mean in an incident. Otter handles alert grouping, severity suggestions, timestamp detection for the timeline and matching similar incidents. Light and fast. In Self-Hosted, it runs inside your environment.
02 Frontier models
Broad knowledge, long context
Drafting reports and postmortems, reading runbooks, and investigating across code and monitoring data use the strongest general-purpose model available at the time. No vendor lock-in: we swap models on accuracy and price.
03 Hybrid
The right model for each job, under your control
The product decides which model handles which job, and admins can narrow the scope. Set, per tenant, what may and may not be sent to external models. Verified Sources behind every suggestion are shown whichever model produced them.
AI proposes. People decide. That design doesn't change.
PRODUCTS
Service operations, starting from the incident.
Every product sits on one record. Start with what you need, and open the next one when you outgrow it.
Incident response
The agent runs first response in your Slack or Teams thread.
Learn more →
Problem, change & release
Trace which change caused the outage and manage known errors.
Learn more →
CMDB & impact map
Connect CIs, business services, sites and suppliers to see what breaks when something stops.
Learn more →
Service desk
Turn requests into tickets. AI classifies, answers first-line, and escalates to an incident when needed.
Learn more →
Tasks & improvements
Chase tasks from incidents, tickets and postmortems until they're done.
Learn more →
Drills
Run exercises from scenarios for 14 industries, with the debrief drafted for you.
Learn more →
Regulatory reporting
Deadlines under GDPR and Japanese breach-notification laws, calculated for you, with drafts in the required format.
Learn more →
Analytics & executive views
Incident volume, recovery times and recurring patterns, for executives, operations and responders.
Learn more →
Works with Datadog, Sentry, Zabbix, ServiceNow, Jira, Backlog, NetBox, GitHub and Notion. Replace them or run alongside. All integrations →
ONE RECORD
More products. Still one record.
Slack conversations, monitoring alerts, code changes, asset configuration and knowledge all flow into the same timeline, and every product reads the same record. A decision in an incident becomes a task, the task becomes a change, the change lands in the CMDB, and that becomes evidence for the next incident.
Incident response
Problem / change / release
CMDB & impact map
Service desk
Tasks & improvements
Drills
Regulatory reporting
Analytics
↑ ↑ ↑
Incident timeline
↑ ↑ ↑
Slack / Teams
Monitoring
GitHub
CMDB
Knowledge
No re-keying
Ticket to incident, incident to report. Nothing gets copied and pasted.
Smarter as it fills up
Similar-incident matching, severity suggestions and recovery-time estimates come from your own past responses.
Audit-ready
Who decided what, and when, can be checked in the audit log and used as evidence for SOC 2 and ISO 27001 audits.
TRUST
You decide how much the AI does.
AI proposes, people decide
Severity, task adoption and resolution are always confirmed by a person. Connections to monitoring, code and documents are read-only. Switch assistance to “Only when needed” or “Off” at any time.
Your data stays put
Self-Hosted runs inside your VPC, Otter included. Set per tenant what may be sent to external models.
Audited operations
Incident Lake has received a SOC 2 Type 1 report, and SIGQ is ISO/IEC 27001 certified. SAML SSO, IP restrictions, audit logs and per-feature permission groups.
Regulatory clocks, counted
GDPR, plus Japan's APPI, My Number Act Art. 29-4 and Active Cyber Defense Act Arts. 4–5. Deadlines calculated from the time of awareness, with drafts in the required format.
Regulatory information here is general in nature and is not legal advice.
The information Incident Lake handles, such as your infrastructure operations data and incident response history, is highly confidential and core to your business. SIGQ Inc., the company behind Incident Lake, is certified to ISO/IEC 27001, the international standard for information security. Incident Lake has also received a SOC 2 Type 1 report, so your data is protected under an internationally recognized security framework.
REPORT
Where incident response stands, in numbers.
72% depend on one or two people. Original research you can take straight into discussions with leadership and front-line teams.
RESEARCH REPORT
Assess where you stand
with a resilience maturity matrix
Resilience Requirements and Security Investment: What Companies Need to Know
Requirements from domestic and international regulations, organized into 8 dimensions × 5 levels. Includes survey data (n=250), case studies and JCIC investment benchmarks.
Learn more →
SURVEY REPORT
250 SRE and engineering leaders on
dependency, causes and risks
Survey on Individual Dependency in Incident Response 2026
72% say incident response depends on one or two people, and 88.9% of them see a risk those people may leave. Eleven questions on the causes and on which countermeasures work.
Learn more →
PRICING
Pay for what you use, not per seat.
Free plan / Free
For individuals and small teams trying it on a first incident.
Monthly AI credit allowance, up to 5 members, 1 Slack / Teams integration, 1 monitoring integration, and up to 20 knowledge items.
Team plan / Monthly
Monthly, for a department. AI credit allowance plus overage within your budget.
Up to 25 members, 20 auto-detection channels, 3 monitoring integrations, 50 knowledge items, and permission groups.
Annual plan / Yearly
For company-wide rollout and regulatory needs. Four tiers: Growth / Scale / Enterprise / Custom, with FDE support.
12-month contract. Annual AI credit allowance, unlimited members, integrations and knowledge items, audit logs, SAML SSO, Public API, and onboarding support.
Capped pay-as-you-go pricing is also available during the PoC.
Contact us anytime for detailed pricing.
FLOW
Getting started
STEP 01
Start
Get your login within 3 business days, often the same day. Then connect Slack or Teams and try it on one incident.
STEP 02
Expand
Import past logs and runbooks, then connect monitoring, CMDB and tickets. An FDE helps you set it up.
STEP 03
Embed
Measure impact with PoC metrics such as recovery time and reporting effort, then roll out company-wide.
FREE TRIAL
Start with one incident in Slack.
Just a work email and company name. We'll email your login within 3 business days, often the same day. No sales calls. For company-wide rollout or Self-Hosted, use the button below.




