Not Chat, But
AI Agents for Managing Personal Data

Create fully customized personal tools like budgeting, weight loss, and more,

Data stored in structured form, not vague chat history

Write, modify, query and summarize data through conversation, making tools useful long-term.

Not chat history, but sustainably actionable data

10K+

Active Agents

100K+

Data Records

5K+

Users

50+

Agent Types

How It Works Example

Case: A Personal Accounting Agent

You want to track income and expenses long-term, and be able to query, summarize and analyze by time, category or notes. But you don't want to maintain complex spreadsheets, nor have data mixed in disorganized chat history.

01
Create Agent

Tell AI this is an accounting Agent. It will create a ledger data table with fields like time, amount, category, and notes.

02
Record via Conversation

Type like chatting: "Lunch today 32 yuan", data will be written to corresponding fields, not staying in the conversation.

03
Query, Modify & Summarize

Ask anytime: "How much did I spend on food this month?" or directly modify any historical record.

04
Generate Statistics

Generate statistical charts by time or category, such as monthly trends or category proportions, for quick analysis.

All records are saved in this Agent's data table

They are not chat history, but sustainably actionable data, which can be repeatedly queried, summarized, and analyzed in chart form.

Creatable Agent Types

You can create different types of Agents for different purposes

Recording & Ledger Agents

Accounting, work logs, asset lists, etc. Maintain structured information long-term, queryable, summarizable and analyzable.

Memo & Knowledge Agents

Memos, meeting notes, reading notes, etc. Content can be organized and traced, not scattered conversations.

Tracking & Analysis Agents

Project progress, habit tracking, metric recording, etc. Data can generate statistical summaries and charts.

Custom Tool Agents

Rule-based checklists, process recorders, state management tools. You define the data structure, Agent executes.

Each Agent operates its own data, not shared chat history.

Core Capabilities Supporting Agents

Making Agents not just an idea, but reusable tools

Agent Marketplace

Structure Reuse, Data Independence

You can get Agents created by others from the Agent marketplace. These Agents reuse structure and rules, each Agent's data is stored independently, no personal data is shared.

Structured Data Display & Visualization

Data Transparency

Each Agent can directly display its internal data tables and records. Data is presented in table and statistical views, not hidden in conversation content.

Scheduled Push & Auto Summary

Continuous Work

You can set scheduled tasks for Agents, such as periodic data summarization, generating statistical results or push reminders, allowing Agents to work continuously even when you're not actively conversing.

Who It's For / Who It's Not For

Finding users who truly need it

Who is this App for?

  • Want to maintain structured data long-term
  • Already using Notion, Excel and other tools to build systems
  • Want to operate data through conversation, not maintain spreadsheets
  • Care about data structure, controllability and reusability

May not be for you, if you...

  • Just want to casually chat with AI
  • Want AI to automatically remember everything, without caring about data structure
  • Only need one-time answers
  • Don't want to spend time thinking about "how to organize data"

FAQ

Clarifying concepts, answering questions

How is this different from regular AI chat?

Conversation is just the operation method, what's actually saved is structured data, not chat history.

What's the difference from "AI with long-term memory"?

You explicitly decide what data each Agent stores and how to store it, rather than AI deciding on its own.

How is this different from Notion, Excel?

You can define and operate data structures through conversation, without directly maintaining tables or formulas.

Will data from different Agents mix together?

No. Each Agent operates its own data table, data is independent of each other.

Do I need to design the data structure beforehand?

No. You can start with natural language, Agent will create initial structure, and can be adjusted anytime later.