AN AI THAT LEARNS YOUR WORK. ON YOUR TERMS.
Build a professional profile, use it in your live meetings and briefs, then keep every learned fact auditable, source-linked, and under your control.
TWIN
Mirrors your work, language, and priorities
LOCAL MEMORY
LEARNING ON
YOUR PROFESSIONAL OPERATING PROFILE
AI & AUTOMATION ARCHITECT
Direct, evidence-led, and focused on decisions that survive beyond the meeting.
FROM PROFILE TO PARTNER
Author Your Profile
Build the professional profile your Twin should work from: manually, with an optional local-AI interview draft, or through a guided deep-creation flow you review before saving.
Meet With Context
Your Twin supports the Live Assistant, live triage, and briefs with the vocabulary, projects, people, and preferences you choose to include.
Review What It Learns
After sessions, the Twin can distill durable facts about people, projects, preferences, and commitments. Every fact stays linked to the session that created it.
Stay in Control
Open the memory map and see the whole picture: a branch per category, every fact on its own line. Read one in full, check the session that taught it, or forget it with a few seconds to undo. Pause learning with one switch, and nothing is added to the profile until you confirm it.
WHAT YOUR TWIN CAN DO
Context makes AI useful. LifeDash brings the right profile into your sessions without turning your private history into an opaque black box.
A profile you own
The Twin begins from a profile you author and review. It is deliberately budgeted per task so it helps without overwhelming the actual session context.
Live assistance that proposes
During a meeting, the Live Assistant can surface proposals and board work. You decide what to accept, and it works visibly alongside the conversation.
Memory with provenance
The memory map opens as structure: a branch per category with an honest count, and every fact on its own line beside the session that taught it. Read it, forget it with a few seconds to undo, or pause the learning loop whenever you choose.
Private by default
The built-in local AI runtime does the work on your own device with no key, no second app, and no account. Point a task at LM Studio or Ollama if you would rather bring your own, or at OpenAI, Anthropic, Google Gemini, or Kimi for the specific task where you want a cloud model.
SEE EXACTLY WHAT YOUR TWIN KNOWS
Memory is only auditable if you can actually look at it. The map opens your Twin as a tree you can explore, branch by branch and fact by fact.
TWIN
28 FACTS
Maya owns the pricing model
Dev team ships on Thursdays
Atlas pilot targets a Q3 launch
Beacon is paused until budget lands
- ›Memory opens as structure. Your Twin sits on the left with one branch per category for people, projects, preferences, and commitments, each carrying an honest count. Multiple branches can stay open at once
- ›Every fact sits on its own line with a short title. Click it to read the fact in full, see the source session that taught it, and forget it, with a few seconds to undo
- ›Point at a fact with the mouse or the keyboard and its category stays readable while the others fade back. Everything stays reachable while faded
- ›Watch it learn during a meeting. A new fact grows into place on its branch, and a closed branch bumps its count instead of yanking the view open
- ›Reduced motion is supported. Everything updates instantly with no animation, and the fading still works
DELETION THAT REACHES THE MEMORY
DELETE A MEETING, DELETE WHAT IT TAUGHT
By design, deleting a meeting deletes what it taught your Twin, and it removes the recording from disk. Before anything goes, a confirmation shows you exactly what will be removed: the fact count with a preview, the recording file size, and whether a brief and transcript exist. Nothing is deleted until you confirm.
If you want the knowledge without the recording, “Keep what the twin learned” is right there as a checkbox, off by default. Kept facts are stamped with a provenance note naming the meeting and the date it was deleted, so their origin never quietly disappears.
Board cards you already accepted from that meeting survive its deletion, by design. The work you committed to does not vanish with the recording that prompted it.
Ask anything across your work…|
↵ ASKCLICK THE FIELD TO ASK YOUR WORKSPACE
YOU OWN THE CONNECTION
LifeDash works out of the box on its built-in local AI runtime, with no key, no second app, and no account. Bring your own key or your own local runtime per task whenever you want to, and cloud use stays explicit, so you know which task uses which provider before content leaves your device.
- ›Choose a provider per task. Live Assistant, Twin Interview Assist, transcription, and embedding can use different models
- ›Local reasoning and local embeddings run on the built-in local AI runtime, with no key, no second app, and no account
- ›LM Studio and Ollama stay available for local models when you would rather bring your own
- ›Cloud use is clearly labelled where content would leave your device
- ›Keys are kept locally with OS-level secure storage
SUPPORTED PROVIDERS
READY TO TRY IT?
Download LifeDash free and see it in action on your own meetings and projects.
Back to all features