SocialOS VC Deck ready 2026-03-09T17:00:34.418Z

Slide 1 · Problem

People, context, content, and self-understanding drift apart.

High-context people do not struggle because they lack tools. They struggle because relationships, conversations, follow-up, content, and reflection fracture across chats, notes, screenshots, and memory.

Meet someone important
Promise to follow up
Turn it into content
Forget what it meant two days later
Chatsimportant details disappear into threads
Notescontext lives in fragments
Contentexpression is disconnected from relationships
Reflectionself-insight rarely becomes action

Slide 2 · What SocialOS is

A local-first relationship and identity operating system.

SocialOS turns messy daily input into structured people memory, event context, platform-native drafts, and daily or weekly mirror loops.

  • One conversational workspace
  • Structured people and event memory
  • Content handoff instead of scattered follow-up
  • Reflection that stays grounded in evidence
Workspace capture view

Slide 3 · Who it is for

Built for high-context people, and already grounded in a real network.

The wedge is not “everyone.” It starts with people who constantly turn conversations into opportunities, follow-up, and public expression. The demo now carries a real relationship graph across London, Bristol, Chengdu, and San Francisco.

London builders Minghan Xiao · Shafi Maahe Imperial hackathon organisers and builder-community operators.
China and legal bridge Candice Tang Cross-border and intellectual-property counsel from Chengdu.
Industry node James Wu NVIDIA leadership and Tianjin University alumni connection from San Francisco.
Bristol teaching Daniel D'Andrea · Alan Champneys · Matt Hennessy Teaching and modelling collaborators across Bristol courses.
Research and enterprise James Sibson · Clare Rees-Zimmerman · Xiyue Zhang · Michele Barbour · Stefan Dienstag Research-industry, workshop, enterprise, and Bristol AI community links.

Slide 4 · Product loop

One real note becomes memory, drafts, queue, and reflection.

This is the core proof: one natural note about a real relationship already moves through the whole product loop instead of dying inside a chat thread.

Real network Minghan Xiao · Candice Tang · James Wu · Xiyue Zhang
Real output contacts · linked event · 7 drafts · queue · mirror

Loop: capture → recall → express → hand off → reflect

Capture in Workspace “Met Minghan Xiao in the London hackathon organiser circle at Imperial College...”

One natural note goes in. No CRM form. No manual copy-paste across tools.

Recall Minghan Xiao + Shafi Maahe + London organiser follow-up
Imperial Tianjin X profile LinkedIn

The same input becomes contact memory, linked identities, and a reusable event thread.

Express + hand off 7 platform-native drafts, then a trust-first queue
LinkedIn X Instagram Zhihu Rednote WeChat

English and Chinese outputs are prepared from the same event, with dry-run queueing by default.

Reflect Mirror closes the loop with next-action judgment
Follow up with Xiyue Zhang Candice Tang partner thread Bristol teaching circle

The system ends with a recommendation and evidence-backed reflection, not a dead note.

Slide 5 · What works today

One real note becomes memory, 7 drafts, and a safe queue.

One seeded relationship note already fans out into contacts, drafts, and a human approval step.

01 · Contacts memory Real named contacts already live in the graph.
02 · Draft generation One London organiser thread already produces 7 drafts.
03 · Approval queue Dry-run and approval stay visible before anything posts.
Input Real note → 1 Contacts → 2 7 drafts → 3 Approval queue
01 Contacts memory
Minghan Xiao, Candice Tang, James Wu

Already visible with notes, tags, and next follow-up timing.

Contacts with real named network
X + LinkedIn follow-up date linked event

This is already real people memory, not a placeholder list or CRM mock.

02 Draft generation
7 drafts from one event

The same relationship thread already fans out across channels.

Draft generation flow
LinkedIn X Instagram Zhihu
03 Approval queue
Dry-run before publish

Human approval stays explicit before anything posts.

Queue and recall flow
dry-run manual approval

Slide 6 · Why it is different

It combines Relationship OS, Content OS, and Self OS in one loop.

Most products optimize one layer. SocialOS connects the full cycle between who matters, what happened, what should be expressed, and what that says about the user.

Relationship OSpeople memory, follow-up, linked events
Content OSplatform-native drafts and queue handoff
Self OSdaily and weekly mirror grounded in evidence
Not just CRM Not just AI writer Not just journaling

Slide 7 · Why this is credible

Trust-first product, real proof, expandable architecture.

The multi-agent layer matters because capture, people memory, reflection, validation, and publishing are different jobs. The user still experiences one calm product surface.

Public repo live
Validation 2026-03-09T17:00:34.418Z
Repo head 49765e3
Product posture local-first · safe rehearsal
Current status Local demo services are healthy and ready for rehearsal.
Public repo Build log and source github.com/zhouyi-xiaoxiao/openclaw-foundry-socialos
Trust boundarylocal-first, loopback-only, safe by default
Capturemodel-first understanding with structured review flows
Linkingpeople and events connected through graph-backed relationships

Slide 8 · What I want now

Design partners and intros for the next unlock.

This is already a working loop. The next unlock is real-data onboarding and low-friction daily use.

What I want nowDesign partners and intros to high-context users who feel this pain today.
What comes nextImport Inbox, multi-entity capture, and LinkedIn mention suggestions.

Ask: If this resonates, I want to talk to people who live across relationships, notes, and content every day.

SocialOS

A local-first relationship and identity operating system for high-context people.

Open zhouyixiaoxiao.org View public repo

Slide 9 · Claw for Human

Claw for Human

Bring Claw into a human-readable relationship workspace.

  • Use /demo for the local recording and /hackathon/#bounty-claw-for-human as the canonical public proof page.
  • OpenClaw Runtime · Workspace UI · Pitch Deck
  • /hackathon/#bounty-claw-for-human
Status ready
Record /demo
Proof claw-for-human

Slide 10 · Animoca Bounty

Animoca Bounty

Persistent identity, memory, and agent coordination for creator/community ops.

  • Use /hackathon?bounty=animoca for recording and the matching /hackathon/#bounty-animoca section for the public proof.
  • OpenClaw Runtime · People Memory · Studio Agents
  • /hackathon/#bounty-animoca
Status ready
Record /hackathon?bounty=animoca
Proof animoca

Slide 11 · Human for Claw

Human for Claw

Friendship and gratitude coaching with visible guardrails.

  • Use /buddy for recording and /hackathon/#bounty-human-for-claw as the canonical public proof page.
  • Buddy Guardrails · People Memory · Self Mirror
  • /hackathon/#bounty-human-for-claw
Status ready
Record /buddy
Proof human-for-claw

Slide 12 · Z.AI General

Z.AI General

GLM inside multilingual Workspace and draft generation, not a side widget.

  • Use /hackathon?bounty=z-ai-general for recording and the public proof JSON to show live GLM integration.
  • GLM Router · Workspace Chat · Draft Generation
  • /hackathon/#bounty-z-ai-general
Status partial
Record /hackathon?bounty=z-ai-general
Proof z-ai-general

Slide 13 · AI Agents for Good

AI Agents for Good

Impact workflows with SDG triage, long-term relationship memory, and multi-channel follow-through via Telegram.

  • Use /hackathon?bounty=ai-agents-for-good for recording and the public proof JSON to show live FLock integration.
  • FLock SDG Triage · OpenClaw Runtime · Events + Drafts · Telegram Volunteer Channel
  • /hackathon/#bounty-ai-agents-for-good
Status partial
Record /hackathon?bounty=ai-agents-for-good
Proof ai-agents-for-good