From connected to resolved
Getting started takes minutes. Nami plugs into the tools you already use, then investigates beside your team — searching sources, reasoning over evidence, and always leaving a concrete next move.
Get started
Connect, set up teams, investigate
Four steps from empty org to your first investigation thread.
- Step 01
Connect your sources
Link your runbook provider, code repository, and log platform. Nami indexes your knowledge base and code so she can search during investigations.
- Step 02
Set up teams
Create as many teams as you need and set how each one hears from Nami. Each team sees only their tools.
- Step 03
Start investigating
Your team asks Nami questions like they would on Slack. She searches, reasons, and responds — adapting her style to each team.
- Step 04
Learn and improve
Nami spots knowledge gaps and drafts the missing SOP or tool for you to review. Every investigation makes the whole system smarter.
Investigation method
Every turn must move the investigation
If Nami cannot give resolution steps or a targeted question, she is not investigating — she is stalling. That failure mode is treated as a product bug, not an acceptable “I don’t have enough information” dead-end.
Search your sources
Nami looks across connected runbooks, code, and logs for evidence relevant to this incident — not a generic internet answer.
Reason over the evidence
She forms and revises hypotheses from what the sources actually say. If two signals disagree, that mismatch becomes part of the investigation — not something to paper over.
Propose a concrete next move
Either specific resolution steps you can execute now, or one targeted question — the missing input that would distinguish the remaining paths.
Stay collaborative
You remain the actor on production changes. Nami investigates with you. Guardrails constrain access and style; they do not replace reasoning.
Checkpoints
Questions that only make sense for this incident
Checkpoint questions are reasoned from the current investigation state — not keyword-matched templates. Each one should be explainable: why this input, and what Nami will do with the answer.
When she needs a time window
“Provide the incident time window so I can isolate the failing log events.”
Logs are required, but the scope is unknown — a vague “more context” ask would waste turns.
When she needs a trace id
“Give me one failed request_id so I can trace the workflow.”
A specific trace would confirm or refute the leading hypothesis.
When two fix paths diverge
“Run this query and paste the result so I can decide whether to backfill or roll back.”
The next action depends on a fact only your environment can provide.
The investigation loop
Every incident makes the next one easier
Nami doesn't just answer — she captures what worked and turns it into reusable knowledge for your whole org.
Investigate
A team member asks Nami a question. She searches your sources, runs relevant tools, and reasons to an answer.
Resolve
Nami suggests fixes and drafts escalation tickets from your templates, so the right people act with full context.
Learn
When Nami spots a knowledge gap, she drafts an SOP or tool. Your next investigation starts a step ahead.
Collaborative, not autonomous
Why we refuse the black-box remediator
Autonomous systems optimize for finishing without you. Collaborative systems optimize for correct progress with you. In production support, the blast radius of a wrong change is usually worse than the cost of one clarifying question.
Nami does not invent resource names or pretend certainty she does not have. Incomplete evidence stays visible. The feeling we design for: “I’m investigating with you, and here’s exactly what I need next to keep moving” — not “Here’s a report. Good luck.”
For product facts and company mission, see About. For plans and packs, see Pricing.
FAQ
Investigation and getting started
What is NamiFlow?+
NamiFlow is an AI investigation navigator. Nami investigates with your team across connected runbooks, code, and logs, reasons over the evidence, and always leaves a concrete next move — resolution steps you can take now, or one targeted question that unblocks the investigation.
Is NamiFlow an autonomous AI agent?+
No. NamiFlow is deliberately collaborative, not autonomous. You remain the actor on production changes. Deterministic rules act as guardrails (access boundaries, team communication style, safety). The investigation itself is AI reasoning over your evidence — not silent remediation.
How does Nami decide what to ask next?+
Checkpoint questions are reasoned from the current investigation state — for example an incident time window to scope logs, or a failed request_id to trace a workflow. They are not keyword-matched templates. Each question should be explainable: why that input, and what Nami will do with the answer.
Which systems can Nami connect to today?+
Live connectors today: Confluence (runbooks/docs), Google Drive (shared folders), GitHub (code), CloudWatch (logs), Datadog (log search), and Jira (escalation create + optional comment writeback). You choose which to connect. Other providers shown as coming soon in Connections are not claimed as available yet.
What access does Nami need, and what can she change?+
Read access by default. Confluence, Google Drive, GitHub, and Jira connect over OAuth (Drive folder access is a Viewer share to NamiFlow’s reader service account); CloudWatch through a cross-account IAM role you create and can revoke; Datadog via API/app keys you paste. You choose which sources to connect and can disconnect any of them whenever you want. Nami does not deploy code, commit code, edit runbooks, modify logs, or make production changes. Writes exist only for Jira escalation: creating an issue and posting updates/comments to a linked issue — each requires a person to click, never automatic silent writes.
How is NamiFlow priced?+
Starter is $20 per user per month. Pro is $49 per user per month. Team plans are Team 15 ($599/mo), Team 40 ($1,499/mo), and Team Unlimited ($2,999/mo) — contact us; there is no self-serve Team checkout. Enterprise is custom — contact us. Investigation packs are 50 for $150. Included investigations are pooled across your team, not siloed per person. New organizations get a 14-day free trial on Starter or Pro (card required). See the live pricing page for volumes and plan details.
What counts as an investigation?+
An investigation is a thread where Nami actually worked the problem — you describe an incident, she searches your sources and investigates with you. Threads you cancel before Nami begins investigating (for example, while she's still asking clarifying questions) don't count. Once underway it counts whether or not you mark it resolved — and if her clarifying questions alone helped you resolve it, that counts too. Investigation counts are approximate: exceptionally large investigations — a small minority with unusually heavy research volume — may count as more than one. Your plan's included investigations are a shared pool across your whole team, not a per-person allowance.
What happens when we use all our included investigations?+
New investigations pause until you add capacity. An org admin can buy investigation packs — 50 additional investigations for $150 — from Settings → Billing. Optionally enable auto-refill there so a pack is charged to your saved card when the pool runs out; auto-refill is off by default and never charges without that opt-in.
Which sources can Nami connect to?+
Out of the box today: Confluence, Google Drive, GitHub, CloudWatch, Datadog, and Jira. She indexes connected knowledge, code, and logs so a single question can search across them. Google Drive is shared-folder access (Viewer to NamiFlow’s reader). Jira is used for escalation create and optional comment writeback on linked issues. Connectors marked coming soon in the product (for example Notion or GitLab) are not available yet.
What access does Nami need to my systems?+
Read access by default. Confluence, Google Drive, GitHub, and Jira connect over OAuth (Drive folders via Viewer share to NamiFlow’s reader); CloudWatch through a cross-account IAM role you create and can revoke; Datadog via API/app keys. You pick which sources to connect and can disconnect any of them at any time. Nami does not deploy code, commit code, edit runbooks, modify logs, or change production. The only writes are Jira escalation actions — create an issue or post an update/comment on a linked issue — and each happens only when a person clicks.
Is my data secure?+
Access is role-based and scoped per team, so members only see the tools and sources they're assigned. Connector credentials are encrypted, and all data is encrypted at rest and in transit. Admins control members, teams, and connected sources from a single dashboard.
Can I try NamiFlow before committing?+
Yes. New organizations get a 14-day free trial on Starter or Pro. A credit card is required at signup; you won't be charged until the trial ends. Cancel anytime during the trial and you won't be billed. Existing organizations that already subscribed are not eligible for another trial.
What can Nami do with Jira today?+
When an investigation escalates, Nami can draft a ticket from your templates. With Jira connected, a person can create the issue in Jira in one click and later post investigation updates as comments on that linked issue. When lifecycle comments are enabled for your org, Nami can also post a brief comment when an escalated investigation is resolved or cancelled. Nami does not change production systems through Jira beyond these ticket writes.
How do teams and access work?+
Admins create as many teams as they need and assign members, tools, and a communication style to each. Each team sees only the tools they're given, and Nami answers in that team's style. Investigations can hand off across teams via investigation ID so context isn't lost.
What is the Founding Partner Program?+
A paid 8-week engagement — $5,000, or $2,500 for small teams — where you run Nami against real production incidents with weekly support from the NamiFlow team. Invoiced upfront, and 100% of the fee is credited toward a 12-month contract if you convert. You pay the annual price and get fourteen months instead of twelve. Limited to three teams at a time and reviewed personally — apply via the contact page. A demo, a workflow discussion, and a historical/sanitized test case are always free; only the 8-week engagement against real environments is paid.
Ready to resolve incidents faster?
Start a 14-day free trial with your team and see Nami investigate on your real production stack.