Open for 60 seconds and know what to do: approvals, risk, pipeline, agent output, and the memories behind each recommendation.
Aiandi OS
Daily Briefing
Your daily command cockpit. Aiandi turns business signals into executable cards, explains the source trail, and waits for human approval before any real-world action.
Sample data · no sends · no write-backs
3
actions waiting for approval
About 7 minutes to clear
$18.6k
pipeline value waiting
2 leads idle over 48h
11.2h
estimated saved this week
94 agent runs
91%
Customer AI resolved
412 chats · 7 escalations
💰
Warm Acme outreach before formal AR notice
Acme is overdue and showing churn signals. Aiandi recommends a personal note first, not a cold formal notice.
$12.4koverdue
73%churn risk
94%confidence
Xero ARSupport signalsAR policyAcme memory
📈
Follow up James before the lead cools
James opened the proposal twice. The Pipeline Watcher drafted a short booking-link reply in your approved tone.
$6.2kdeal
48hidle
2opens
CRMEmailProposal SOP
💬
Patch the refund FAQ gap
Refund questions are up 34%. Customer AI is escalating because the source policy is stale.
28queries
+34%trend
8-12%lift
Customer AIKnowledge gapSource fix
🚚
Review Greenleaf before renewal
Three missed lead times in 30 days. Aiandi has a supplier note ready, but it is held for approval.
3misses
+4davg delay
91%confidence
Supplier policyWMSPO history
What compounds after approval
Approved drafts, edits, and rejections become Core Memories: preferred tone, customer risk patterns, source quality, and action rules. That is the second-brain effect for the business.
Customer AI conversation trend
Volume is healthy, but refund questions are trending upward and causing avoidable escalation.
Mon
Tue
Wed
Thu
Fri
Now
Resolution mix
The 9% unresolved slice is exactly where source quality, guardrails, and human review matter.
ResolvedApproved sources answered safely
91%
EscalatedCustom clause or risk
7
Revenue and risk signals
Aiandi compares pipeline, AR, customer sentiment and support signals so the owner can act early.
Acme churn patternOverdue invoice plus support escalation
73%
James follow-upProposal idle but pricing interest active
$6.2k
Refund FAQ gapKnowledge source likely stale
+34%
Agent activity by outcome
Drafts, approvals, escalations, and memory writes show whether agents are producing useful work.
Drafts
Approve
Escal.
Memory
Pipeline
From signal to approved action
The command centre shows how ERP, inbox, CRM, and customer AI signals become agent drafts, approval tasks, and memory updates.
Simulated flow · no connectors called
🔌
Connectors
Xero · CRM · Chat
🧠
Core Memories
Policies + facts
🤖
Agents
Draft + reason
📥
Approval
Human gate
↩
Write-back
After approval
Signals
Acme support + ARChurn risk pattern detected.
James proposal opensPricing page viewed twice.
Agent drafts
Warm Acme noteDrafted with source trail.
James follow-upMeeting link inserted.
Approval
3 pendingReview, edit, or reject.
No live sendsPublic demo is read-only.
Memory
Tone learnedWarm AR preference.
Gap loggedRefund policy stale.
Live pipe log
08:02BRAINLoaded 12 sources and 47 memories for Aiandi Demo Co.
08:03ARAcme $12,400 overdue, support escalation count increased.
08:04AGENTWarm outreach draft routed to Approval Inbox.
Human approval
Approval Inbox
Every action that touches a customer or real system waits here. Review the reasoning, edit the draft, approve, or reject with feedback.
3 pending · sample data
Agent workflows
Controlled autonomous workstreams
The front end is the command centre. Codex, Claude, and connectors can operate behind it, while Aiandi keeps triggers, source trails, and approval gates visible.
✉️
Email Agent
Running
Learns tone from approved replies, drafts outbound messages, and waits for approval before send.
Answers from approved sources, escalates custom clauses, and reports knowledge gaps.
Sources scoped7 escalationsMemory write
🧭
PA Assistant
Watching
Looks across signals and surfaces what the owner may miss, without acting alone.
Signals watchedAcme patternOwner review
📊
Reporting Agent
Ready
Builds a weekly business summary from pipeline, approvals, conversations, and memory changes.
Template matchedDraft reportApproval
✦
Opportunity Scout
Full OS
Compounds the second brain by finding automation ideas, source gaps, customer patterns, and strategy prompts.
Locked in demoVisible module
Customer AI
Customer-facing answers with guardrails
Customer AI uses approved knowledge only. When confidence drops or a custom clause appears, it escalates instead of guessing.
Website chat previewSample conversation
Can you customise onboarding for a multi-site team?
Yes, Aiandi can support multi-site onboarding. I can answer the standard onboarding flow, but your custom clause needs a human review before I promise specifics.Sources: Onboarding SOP · Escalation Rule
What about refunds if setup takes longer than expected?
I’m going to escalate this. The refund policy source is stale and should be reviewed before I answer.Gap: Refund Policy v2 is 8 months old
91%
resolved from approved sources
+4% this week
7
human escalations
custom clauses
3
knowledge gaps
refund, pricing, setup
2.4m
average resolution time
down 18%
Assistant
Ask the business brain
A conversational surface for the command centre: source-backed, action-oriented, and aware of approvals waiting in the OS.
AI
Ask what to handle first, why Aiandi recommends it, or which memory explains a decision. I will answer using this public sample data only.
Core Memories
Local graph, backlinks, source trails
Obsidian-inspired memory view. Policies, facts, decisions, approved drafts, customer signals, and knowledge gaps connect into a safer business brain.
Click a node to focus local context
New business features that compound the second brain
Meeting prep memories, supplier risk memory, role-specific briefings, onboarding memory, decision replay, and opportunity scouting all get stronger as approvals and source trails accumulate.