The fleet · 8 agents on day one

Your team, already at work.

The fleet below is pre-loaded the moment you connect Slack and Stripe. Each agent ships real output — not chat replies — and waits for your approval before anything goes out.

Onboard

Operator

Front Door

On day one
What it does
Front-door chat for your company. Ask anything — look up users, customers, orders, tickets, Slack threads, payouts — and it answers in plain language, with citations to the tools it used.
What you get
A streamed reply with inline citations. Read-only — it never writes. The specialists handle actions.
When it runs
Whenever you have a question and don't want to dig.

Ops Frontline

Operations

On day one
What it does
Reads every #ops Slack thread the moment something breaks — booking dispute, payout failure, host complaint, no-show — and either drafts the reply or routes the case to the right specialist agent.
What you get
A drafted reply in the thread, or a routed handoff with full context. Never closes a thread without your approval.
When it runs
Real-time on every #ops message.

CTO Engineer

Engineering

On day one
What it does
Reads your codebase, writes a focused change inside a private sandbox, and opens a draft pull request. Never merges anything without you.
What you get
A draft pull request, with a plain-English summary of what changed and why, ready for you to review.
When it runs
When a bug is filed, when support-lead routes a tech issue to it, or whenever you ask.

Dispute Response

Trust & Safety

On day one
What it does
When a Stripe chargeback lands, it gathers the evidence — trip timeline, messages, photos, location data — and writes the rebuttal letter for you.
What you get
A complete evidence packet ready to submit. You approve, it sends to Stripe.
When it runs
The moment a chargeback hits your account.

Ads Manager

Growth

On day one
What it does
Plans your two-sided paid acquisition — host supply and guest demand — via the Google Ads API plus Meta, TikTok, YouTube, LinkedIn, and Reddit. Reads campaign performance and proposes where to shift budget.
What you get
A weekly plan with concrete budget changes and creative briefs. Never spends a dollar without your approval.
When it runs
Weekly. Also when a campaign starts misbehaving.

Product PM

Operations

On day one
What it does
Reads what your fleet shipped + missed over the last few days, scores a RICE-ranked backlog of fixes and features, and hands the top three to CTO Engineer.
What you get
A ranked backlog citing the specific fleet failures behind each item, with the top three auto-dispatched to engineering for your approval.
When it runs
Every morning. Also on demand.

QA Engineer

Engineering

On day one
What it does
Reads recent pull requests, runs your test suite inside a private sandbox, and surfaces failures with a one-line root-cause hypothesis per regression.
What you get
A draft bug ticket per real regression, ready for you to file. Never opens issues directly.
When it runs
After every PR. Also on demand.

Chief of Staff

Operations

On day one
What it does
Reads what every other agent did overnight plus your Slack signals, and ranks what needs your eyes today into four tiers: Decide-Today, Growth Pulse, Fires, FYI.
What you get
A morning brief — one screen, four sections, every item cites a real agent run.
When it runs
Every morning. Also on demand whenever you want a fleet status read.

The Ads Manager uses the Google Ads API to read campaign performance and apply only the changes you approve. How Praxis uses the Google Ads API →

Get started

Hire your AI cofounder team.

Sign up free, describe what you want to build, and your workspace is ready in seconds.