Opportunities › Automated root cause analysis for on-call incidents › Proposed app
Firstcause
An on-call agent that follows your runbook and shows its evidence
- Score
- 47
- confidence 69
- Evidence
- 17
- items on the page
- Rivals
- 6
- named in sources
- First version
- 45 d
- estimate
- Type
- Desktop app
- PWA: runs in the browser, installs on Mac and Windows
- Market since
- 2023
- Emerging market · AI estimate
ScreensConcept
Desktop app: On-call engineers analyze runbooks, telemetry data and evidence-rich root causes during incidents at a desk. Sample screens on a computer, in the proposed brand. Their numbers are sample data.
Incident 482
Investigating, read-only3 of 6 steps
About this appAI proposal
Firstcause runs your on-call runbook when an alert fires. It follows fixed investigation steps: isolate the failing service, check the traffic surge, scan recent deploys, search log patterns. It then posts a ranked shortlist of likely root causes in your incident chat channel.
It is for engineers on call at companies that run production systems and already use observability tools. Each conclusion links to the queries, spans and log groups behind it, so you can verify it quickly instead of trusting a guess.
It differs from pasting logs into a general chat model. Large telemetry results are saved outside the model's context, tools are read-only by default, and every step follows a runbook your team wrote and approved.
- Tagline
- Follow the runbook. Show the evidence. Find the first cause.
- Voice
- CalmPrecisePlainCandid
- Why this name
- The first cause is what an on-call engineer wants at 3 a.m. Two plain words, easy to say aloud in an incident call.
In the first version
Members say which features they would use. It tells PWAS what people want; it changes no score.
Runbooks as ordered steps
Write investigation steps once. The agent follows them in order and records which step produced which finding.
No member yet
Telemetry tools built for the job
Anomaly detection on metrics, log pattern search and span tree reasoning, instead of raw logs pasted into a prompt.
No member yet
Evidence beside every cause
Each suspected cause lists the queries, spans and log groups behind it, plus what the agent could not check.
No member yet
Big results stay out of context
Large telemetry outputs are stored on disk and read in slices, which limits context overflow and wasted tokens.
No member yet
Brand identityAI proposal
Check the name against trademark registers before using it.
Palette
#0F5E57Deep Teal · primary#F6F3ECPaper · surface#1B2B2AInk · text#E0743AEmber · accent#3D6FA3Slate Blue · data#B3382CBrick · alert
Typography
Aa
Firstcause
PricingAI proposal
The material gives no on-call tool prices. The only verified competitor price is Cursor: $20 to $40 a month, with a free plan. The tiers here are a starting point to test, not a market claim.
Trial
Free
- 1 runbook
- 1 observability connection
- 10 investigations a month
Team
$99 a month
- Up to 5 on-call engineers
- Unlimited runbooks
- 3 observability connections
- Chat channel delivery
Business
$299 a month
- Up to 20 on-call engineers
- Approval and audit log
- Self-hosted agent option
- Priority support
What it could be worthEstimate
What a software business making this much sells for today, from real deals. Pick a monthly revenue.
If it makes
Worth about
$194,000
Sale range $132,000 to $263,000
$120,000 a year in revenue
From 87 software businesses listed and sold on Empire Flippers in the last year, revenue vetted by the broker: they sell for about 19× their monthly revenue, most between 13× and 26×.
Measured on profit, the middle deal is 32× monthly net profit: about 2.7 years of profit.
An estimate, not an offer or financial advice. A real price depends on profit, growth, churn, how much of the work is the owner's, and the buyer.
How it would growAI proposal
How the product would reach new customers through its own use, and how the first ones would be found. A plan to test, not a result.
- Built-in loop
- Shared outputs. Each investigation produces a timeline link posted in the incident thread. Teammates, managers and other teams open it without an account and see the evidence and the product's name.
- Do they become customers?
- Readers are often other on-call engineers who could set up their own runbook. Test how many viewers click through to start a trial, and whether a team lead then buys.
- First customers
- Platform or SRE teams of 5 to 30 on-call engineers at SaaS companies with large telemetry volumes, using Prometheus and Grafana on Kubernetes.
- Where to reach them
- Hacker News and GitHub: Post a technical write-up of runbook steps and the disk-saved telemetry approach, and answer engineers in comments on related threads.
- SRE communities: Offer free runbook reviews in SRE Slack groups and forums, then run the agent on a past incident.
- Direct outreach: Message SRE leads who posted about on-call burden, offering a trial on their recorded incidents.
- First milestone
- Five teams run it on live alerts for a month, and at least two pay for the Team plan.
Why people already pay for this
- VerifiedHacker News post received 48 points and 25 comments
- AI inferenceTeams are using general AI tools to help reduce on-call burden by pasting logs into AI and using Claude with observability integrations.
- AI inferenceAutonomous root cause analysis is a difficult problem for AI, with current state-of-the-art models achieving poor accuracy on benchmark tasks.
- Public claimOn-call investigations have three core technical challenges: telemetry data noise, context-dependent interpretation, and high-stakes time constraints where errors are costly.
Tech stack
Every proposal builds on the same PWA base, installable from the browser on any phone or computer. Dashed boxes are what this app adds.
- AppNext.js and TypeScript
- PWAManifest and service worker
- UITailwind CSS
- Data and authSupabase Postgres
- PaymentsStripe
- EmailResend
- HostingVercel
- This app · Agent runtimeSandboxed worker service with read-only tool permissionsPermissions must be enforced in code and the agent runs longer than a serverless function allows.
- This app · Telemetry connectorsConnectors for Prometheus, Grafana and log or trace storesThe agent needs to query metrics, logs and spans where teams already keep them.
RisksAI inference
- AIAutonomous root cause analysis is hard. The material cites one frontier model at 36% on a benchmark. Wrong causes shown with confidence could mislead engineers during an incident.
- TechnicalEach team's stack, naming and log formats differ. Connectors and runbook tuning may take more per-customer work than a small team can absorb.
- RegulatoryProduction telemetry can contain personal or secret data. Buyers may demand self-hosting, redaction and review before they send it to a hosted model.
- CommercialTeams already use general AI tools with observability integrations, and open-source agents exist. It must beat them on accuracy and trust to justify a subscription.
- PlatformDepends on access to observability and chat vendors' APIs, whose terms, rate limits and data access can change.
Everything marked AI proposal or AI inference was written by a model from the evidence on the opportunity page. The proposed name must be checked against trademark registers before it is used.