Describe the outcome
"Research 20 fintech companies in Germany and draft a first-touch email for each." Add acceptance criteria and a budget.
Insiro Multiplayer
Give AI teammates a goal. Keep the decisions.
Create a team of AI engineers, analysts and admins inside Insiro. Describe the outcome, set a budget and the apps they may touch. The lead teammate builds the team, does the work and asks you only when it needs you.
GoalsLead + sub-agentsAcceptance criteriaSpending capsLive steering
Goal 2194 · liveBudget $6.40 / $25
MMira Lead teammate
RResearcher
LinkedIn
LinkedInAAnalyst
Snowflake
SnowflakeWWriter
Gmail
GmailVReviewer
Notion
NotionLead reasoning
Goal: 20 fintech leads + drafts Criteria: 3 checks · cap $25 Plan: 4 roles · 14 tasks Risk: email send = tier 2 3/3 criteria met 0 policy blocks $6.40 of $25
Lead reasons about the goal. Mira reads the goal and acceptance criteria, splits the work into four roles and checks the budget before anyone starts.
Work flows to sub-agents. Each sub-agent gets one assignment, a slice of the budget and only the apps its role needs. Nothing wider than the lead's own access.
Sub-agents talk to each other. The researcher hands accounts to the analyst, the analyst passes scores to the writer, and the reviewer flags a gap, all visible in the activity feed.
Results flow back up. Each sub-agent returns its output with evidence. Nothing is accepted on trust: the lead checks it against the criteria.
Verified result reaches you. Mira verifies all criteria, stays inside the $25 cap and hands you the result with a receipt of every tool call.
What is Insiro Multiplayer?
Insiro Multiplayer lets you work alongside AI teammates. You describe an outcome, set a budget and choose which apps may be used. A lead teammate drafts a plan, builds a small team of specialised sub-agents, does the work in those apps and asks you only when it needs a decision or input.
- 1lead teammate per goal
- 4+specialist sub-agents
- $hard spending cap per goal
- 100%criteria checked before handoff
How a goal runs
Four steps from outcome to verified result.
Meet your teammate
Insiro drafts a lead teammate with a role, persona and the apps it may use. Edit anything, or pick an existing teammate.
Approve the plan
The lead proposes sub-agents and tasks. Apps outside its ceiling are removed and shown. Approve, re-plan or discard.
Get a checked result
The team works in your apps, verifies each task and checks the result against every criterion.
Your team, visible
Watch the work move: you → lead → team → lead → you.
You hand the goal to the lead teammate. It splits the work across sub-agents, they deliver inside their app ceilings, and the lead verifies everything before handing the result back to you, inside the budget and guardrails you set.
- Team graph with live status, app badges and sub-agents added mid-run.
- Computer tab: when a sub-agent opens a browser, watch it and take over.
- Result split into sections per teammate, with structured data as tables.
Why Insiro, not the alternativesSingle-agent assistants hide a long job behind one loading state and one wall of text. Multiplayer shows who is doing what, in which app, at what cost, so you can trust the result or step in early.
1. Set the goal. Objective, 3 acceptance criteria, $25 cap, apps allowed: LinkedIn, Snowflake, Gmail.
2. Plans the team. Mira splits the goal into 14 tasks for 4 roles and reserves budget for each.
3. Works in parallel. Sub-agents work inside their app ceiling. Email sends wait for your approval (tier 2).
4. Verifies the result. Every task is checked against the criteria; one gap goes back to the writer and is fixed.
5. Receive the handoff. Result, evidence and an itemised receipt: $6.40 of $25, 0 policy blocks.
Budget used$6.40 / $25
App ceiling Read-only by default Approval for sends PII redaction
Attach a goal
Every agent knows exactly what it is responsible for.
A goal in Insiro is an objective plus acceptance criteria, a budget, a deadline and the apps allowed. Attaching it to a teammate gives every sub-agent a slice of that goal and a role-specific objective, so each one stays focused, can be measured, and declines work that falls outside its brief.
- Role objectives: each sub-agent inherits one clear outcome from the goal, not a vague prompt.
- Acceptance criteria are checked task by task, then for the whole goal.
- Drift is caught: out-of-scope requests go back to the lead instead of being acted on.
- Accountability: every action is traced to the goal, the role and the person who approved it.
Goal 2194 · attached
Research 20 fintech leads in Germany and draft outreach
- 20 accounts, each with a named buyer
- One draft per account in our voice
- No email sent without approval
$25 cap Due Fri 3 apps
R
ResearcherObjective: Find 20 German fintechs and their buyersLinkedIn · Apollo
98%on objective
A
AnalystObjective: Score each account on fit and timingSnowflake · read-only
96%on objective
W
WriterObjective: Draft one first-touch email per accountGmail · drafts only
94%on objective
W
"Also update our pricing page?" Outside goal 2194 · not in the Writer's role · asked Mira instead of acting
You stay in charge
Steer in plain language. Approve what matters.
Message the lead teammate at any time to change direction; it updates the plan and tells you what changed. Clarification requests and held actions wait in one Needs you queue.
- App ceilings: a sub-agent can never go beyond the apps its teammate may use.
- Spending caps per goal, with a clean stop at the limit.
- Exact-action approvals: what you approve is what runs, once.
- Try again re-runs only failed steps and keeps finished work.
Why Insiro, not the alternativesAutonomous-agent demos run wide open or stop for permission at every click. Insiro grants apps per goal up front, so the team moves fast inside the lines you drew and pauses only for decisions you reserved.
M
Mira · Lead teammate
Drafts are ready for 20 companies. Writer wants to send the first 5 from your Gmail.
Writer wants to run: gmail.send_email
to: 5 recipients · subject: "Faster reconciliation for …"
risk: tier 2 · runs exactly once
risk: tier 2 · runs exactly once
Approve this actionReject
Approve. Skip anyone we already talk to in HubSpot.
Start from a role
Teammates for every function.
Each template comes with a persona, a suggested model, a default budget and recommended apps. Make it yours in a minute.
Sales Researcher
Finds and qualifies accounts, maps the buying committee and drafts first-touch outreach in your CRM.
Talent Scout
Sources candidates against a role brief, summarises profiles and prepares a shortlist for the hiring manager.
Paid Media Analyst
Pulls spend and conversion data, spots waste and drafts budget shifts for your approval.
Chief of Staff
Prepares briefs, chases action items across tools and keeps leadership updates on schedule.
Support Triage
Classifies incoming tickets, finds similar resolved cases and drafts replies or escalations.
Engineering Debugger
Reproduces an issue in a sandbox, reads logs and code, and opens a pull request with the fix.
Why Insiro
Multiplayer vs. the alternatives.
| Capability | Insiro Multiplayer | Single AI assistant | Autonomous agent experiments |
|---|---|---|---|
| Works toward an outcome, not a prompt | Goals with acceptance criteria | One reply at a time | Loosely defined |
| Builds its own team | Lead + specialised sub-agents | No | Often loops |
| Acts in your real apps | Within an app ceiling | Limited actions | Rarely enterprise-ready |
| Budget per goal | Hard cap, itemised | No | Runaway costs |
| Human steering mid-run | Message the lead any time | Start over | Rare |
| Verified result | Criteria marked met / not met | No | No |
| Turn into a recurring routine | One click | No | No |
What is Insiro Multiplayer?
Insiro Multiplayer lets you work alongside AI teammates. You describe an outcome, set a budget and choose which apps may be used. A lead teammate drafts a plan, builds a small team of specialised sub-agents, does the work in those apps and asks you only when it needs a decision or input.
What is the difference between a goal, a teammate and a sub-agent?
A goal is the outcome you want, with acceptance criteria and a spending cap. A teammate is a persistent AI colleague with a role, persona, model and an app ceiling. For each goal, the lead teammate creates sub-agents, such as a researcher, analyst or writer, that can never use apps beyond the teammate's ceiling.
Can I create a team of AI engineers, developers or admins?
Yes. Start from role templates such as Sales Researcher, Talent Scout, Paid Media Analyst, Chief of Staff, Support Triage or Engineering Debugger, or define your own. Each teammate has a model, a default budget and the apps it may use. You can pause, edit, archive or give a teammate a new goal at any time.
How much control do I keep over AI teammates?
You approve the plan before work starts, see every task and spend live, and can steer the lead teammate with a message at any time. Apps you grant are used without asking again for that goal; anything held for a decision lands in your Needs you queue, and approving runs exactly what you saw, once.
How does Insiro know a goal is done?
You write acceptance criteria when you create the goal, one per line. Every task's output is verified, and the final result is checked against each criterion and marked met or not met with a reason. If steps fail, Try again re-runs only what failed and keeps finished work.
Can a finished goal become a recurring routine?
Yes. Any completed or partly completed goal can be turned into a routine with one click, so the same team runs it on a schedule, for example a weekly competitor digest or a monthly close checklist, with the same budget, apps and approvals.
Bring a real goal. Watch a team take it on.
In the demo, you describe an outcome from your backlog and we run it with a teammate on your connected apps, inside a budget you set.
- 13governance checks on every agent tool call
- 1,700+SaaS apps agents can act in
- 30+data dialects for plain-English questions