SYSCORTEX v2.4
MODEINTAKE
LATENCY142ms
FIDELITY
CORTEX
SIGN INGET STARTED →
Personal intelligence infrastructure

Build your alternate intelligence.

Documents capture facts. Cortex captures psychology — how you decide, where you won't bend, how you sound under pressure. Then it answers as you in chat, joins meetings, and plugs into any AI.

GET STARTED →PREVIEW ONBOARDING
FIDELITY TARGET ≥70%LAYERS 6DEFAULT PRIVATE
Identity subgraphLive
VALUESVOICEJUDGMENTMEMORYTENSION
NODES
CONFIDENCE
DRIFT
LOW
[01] What would Aashwin say about pricing?[02] Can your twin take the 4pm standup?[03] Ask my Cortex first.[04] What's her honest take on remote work?[05] Send the bot — I'm double-booked.[01] What would Aashwin say about pricing?[02] Can your twin take the 4pm standup?[03] Ask my Cortex first.[04] What's her honest take on remote work?[05] Send the bot — I'm double-booked.
The problem
§01

Dumping files into a vector store is not building intelligence.

Documents capture facts. They don't capture judgment, tension, uncertainty, or how you shift tone by audience. Cortex interviews you psychology-first — then compounds that into a clone others can query.

Protocol

How it works

Phase 0101

Teach it

Not your files. You.

Psychology interviews on how you decide, what you value, where you won't budge. Voice notes and writing add the texture.

Phase 0202

Build it

A living intelligence.

Beliefs, style, judgment calls, and the tensions between them — not a chatbot persona.

Phase 0303

Deploy it

You, everywhere.

Your voice in meetings. Your judgment over REST and MCP. Anywhere you can't be, your Cortex already is.

Meeting agent

Your clone takes the meeting.

Paste a Meet link. Your Cortex joins, listens, and answers in your cloned voice with your actual opinions. The transcript feeds back into its intelligence when it's done.

GOOGLE MEET · SESSION 0042● SPEAKING
ATTENDEE · 14:32:08
“Should we take this deal at 15% equity?”
CORTEX · RESPONSE
“Honestly? No — the math doesn't clear my bar. I'd counter with 12% and a performance kicker.”
CONFIDENCE: 0.91
LATENCY: 840ms
VOICE: CLONED
Attach arms

Build intelligence first.
Then deploy it everywhere.

STARTER+

Public chat + REST

Others query your clone. Any stack calls your judgment via API.

PRO+

MCP + Voice

Claude, Cursor, and custom agents get your reasoning. Spoken replies in your voice.

CONCIERGE+

Meeting agent

Your clone joins Google Meet, listens, and speaks with your actual opinions.

POST /you/api/query
MCP for external agents
REALTIME voice over WebRTC
Pricing

Same clone quality.
Different reach.

Starter
$20/mo

Build your intelligence and share it through chat and REST.

  • Public clone chat
  • REST API
  • 500 queries/mo
GET STARTED →
Recommended
$100/mo

Plug your clone into agents with MCP and spoken voice replies.

  • Public clone chat
  • REST API
  • MCP for agents
  • Voice replies
  • 5,000 queries/mo
GET STARTED →
Concierge
$500/mo

Unlimited reach, meeting agent, human onboarding session.

  • Public clone chat
  • REST API
  • MCP for agents
  • Voice replies
  • Meeting agent
  • Coding agent loop
  • Unlimited queries
GET STARTED →
Vision

There is only one of you. Be everywhere anyway.

GET STARTED →
DESIGN PREVIEW · PSYCH-OS · COMPARE →
v1.246