Ask
Operational questions
Ask “what did I miss this week?”, “check CRM follow-ups”, “who needs attention?” or “what should I approve next?” AEGIS routes those questions to operational sources instead of generic document search.
Recorded walkthrough
The full video goes here.
The video slot stays honest until the recording is complete. It will show the current AEGIS flow, not a staged fantasy workflow.
▶
Full AEGIS walkthrough
Coming after the current public-presence pass. The intended recording: set up a workspace, connect sources, capture notes, review tasks, ask operational questions, inspect catches, approve what becomes memory/action.
Reserved embed area: YouTube/Vimeo/hosted video link will replace this placeholder.
Until then, use the live pages: Guardian overview, Tech Stack, Vigilance Guide, and Evaluator Q&A.
Open Guardian overviewDaily use
What a smaller team actually runs in AEGIS.
The showcase is not about another dashboard. It is about the daily work AEGIS captures, organizes, prepares, remembers and refuses to push through without approval.
Notice
Vigilance catches
Radar, document-policy vigilance and source-backed loops surface pending catches with evidence. The human approves or dismisses before the catch becomes durable memory.
Decide
Human-gated action
AEGIS can prepare drafts and next steps. Sending email, writing externally or preserving trusted memory remains behind the signature.
Remember
Memoria
Memoria is continuously curated institutional memory: evidence-backed, human-approved organizational knowledge, not a raw vector dump.
Stay present
Guardian + Standalone Omni
Guardian is where AEGIS gets your attention. Omni is where you ask, decide, review and approve. `/os` remains the deeper workspace.
Prove
ADE / Proof Fabric
ADE is the repeatable proof kit: synthetic source data, expected catches, real source ingestion, captured outputs and verifier reports with honest scope. v6 is ratified for the fixed eight-category taxonomy.
Daily operations
For one founder, or a 25-person team.
AEGIS is useful before a company has enterprise process. It gives small teams a shared operating rhythm across the places where work already happens.
Ben-style usage
A realistic solo-operator day
- Research briefs arrive through scheduled Telegram routines.
- Email checks are prepared for follow-up instead of buried in the inbox.
- Notes are captured from Telegram or the workspace and kept available later.
- Docs from Google Drive or other work surfaces become searchable through the knowledge base when uploaded or connected.
- Omni and Guardian stay available when the operator is not sitting inside a dashboard.
01Capture: notes, reminders, documents, CRM updates, inbox signals and open tasks from the channels the user already uses.
02Organize: tasks, schedules, workflows, skills and templates turn recurring work into repeatable operating routines.
03Prepare: AEGIS drafts the briefing, follow-up, checklist or next action with source context attached.
04Decide: Boardroom, Omni and approval surfaces keep the human in the chair before external sends or durable memory writes.
05Remember: Memoria preserves only evidence-backed, approved institutional knowledge so the next decision starts with context.
06Notice: Guardian and Catch Ledger surface missed follow-ups, stale commitments, policy discrepancies and other vigilance findings.
Operating flow
The current AEGIS loop.
This is the product spine that matters for SMBs and SMEs: not more chat, but a governed loop from source evidence to human decision and back into daily operations.
Connect sourcesCRM/HubSpot, Google Drive, notes, email paths, tasks, schedules, workflows, calendar, docs and Memoria.
Ask, capture or waitThe user asks in Omni/Guardian, sends a note, uploads a doc, or scheduled loops/digests bring attention to pending work.
Catch the missAEGIS proposes catches with source refs, provider links and enough context for review.
Prepare next stepAccepted catches and daily work items can prepare internal tasks, checklists, schedules or source-native drafts, generation-only until approved.
Human signsNothing consequential is sent, written externally or preserved as trusted memory without the human gate.
Memory compoundsApproved knowledge enters Memoria and becomes future context for agents whose grants allow it.
What to test
Six concrete demo tests.
These are the practical checks a future customer or internal tester should understand before trusting the system.
1Ask “what did I miss?” and inspect whether sources are operational, not random docs.
2Open Catch Ledger and review evidence before approving or dismissing anything.
3Use Standalone Omni or Guardian as the compact daily channel.
4Approve a prepared draft only after checking the source context.
5Create a note from Telegram or the workspace, then confirm it is available for later work.
6Configure a scheduled research/email/task routine and verify the output arrives in the expected channel.
Honest status
What is live, hardening and next.
This page avoids pretending that every future source or proof layer is finished. The current product is already useful; the claim boundary stays explicit.
/os, Standalone Omni, Guardian Windows pilot, Catch Ledger, Memoria, HubSpot read-through evidence and CRM follow-up routing, Telegram vigilance commands, notes/tasks/calendar/docs, schedules, workflows, skills, governed learning suggestions, signed in-system skill/workflow improvements, Boardroom and human-gated prepared drafts.
Broader operational-question grounding, Guardian signing/update path, Chrome extension packaging, richer source rendering, public video, broader ADE proof packs and additional signed improvement appliers.
M365 self-service, CRM import/write-back, Dynamics/Salesforce/SuiteCRM packs, note-source vigilance, broader Proof Fabric packs and claim-grade sovereign verification.
Test the vigilance loop.
Start with a clean workspace, connect the sources you actually use, add one note, one schedule and one CRM/doc source, then ask the broad operational questions and inspect the evidence gates.