Product Definition · v0.3 · Design Studio Ⅱ 2026 Spring

NestAI

栖巢 · 产品定义文档

Status  Co-authored draft Team  Vendow × Wenhuan Version  v0.3
00

Core Thesis

v0.3 核心结论

NestAI is not a conventional AI interior design tool, nor a generator of renderings, floor plans, or shopping lists. Its central proposition is:

Help young renters and dormitory residents grow their own lifestyle inside constrained, borrowed, impermanent spaces.

帮助泛年轻租住人群与宿舍人群,在有限、受约束、未必长期属于自己的居住空间中,生长出独属于自己的生活方式。

Five key shifts mark the evolution from v0.1 / v0.2 to v0.3:

From spatial identity display to lifestyle growth

NestAI doesn't summarise who the user is — it helps them externalise a desired lifestyle through repeated spatial interventions, object selections, moves, adjustments, and feedback.

从"空间人格显化"推进到"生活方式生长"

From design generation to spatial intervention

The emphasis is not a complex report, but a compelling vision of a possible life — then one extremely light, executable action to start.

从"设计方案生成"推进到"空间干预行动"

From anti-template to absorbing templates

Platform content — Xiaohongshu, Pinterest, Instagram — is not the enemy. It is the system's nutrient. The real problem is that external inspiration doesn't automatically fit the user's real person, real space, and real constraints.

从"反模板"推进到"吸收模板"

From cold-start forms to constraints emerging through feedback

User constraints should not be collected via heavy onboarding questionnaires. They are revealed gradually as spatial interventions are proposed and users respond.

从"冷启动填表"推进到"反馈中显影"

From virtual immersion to physical world action

Most software pulls users deeper into a screen. NestAI does the opposite — it uses virtual understanding and simulation to push users back into the physical world to act.

从"虚拟世界沉浸"推进到"物理世界行动"
01

Product Narrative

产品叙事

1.1  Background

背景

Young people across contemporary East Asian and Chinese-diaspora contexts increasingly live in constrained spaces: dormitories, rental flats, long-term serviced apartments, HDB units, micro-apartments. These spaces share a common character — limited area, fixed structure, restricted permissions, uncertain tenure, limited budget, and highly overlapping functions.

Yet these residents are not without aspirations. They scroll through Xiaohongshu, Pinterest, Instagram, Bilibili, Douyin — collecting thousands of room transformations, desk setups, lighting ideas, plant arrangements, storage systems, lifestyle aesthetics.

They do not lack inspiration. What they lack is a system that translates external inspiration into their own real life, real space, and real constraints.

他们不缺灵感。真正的问题是:外部灵感不会自动适配用户真实的人、真实空间和真实约束。

1.2  Core Users

核心用户

NestAI v0.3 defines its primary users as young renters and dormitory residents — people operating under limited budget, limited spatial permissions, and uncertain tenure, yet unwilling to accept a space that merely "works."

Product-scale user

Young renters, dormitory students, early-career renters, city youth in long-term apartments, HDB / transitional housing residents

产品本体用户:泛年轻租住与宿舍人群

Studio MVP user

Tsinghua SIGS Phase II dormitories — 7+1 suite layout — postgraduate living in standardised dorm spaces

Studio MVP:清华深研院二期宿舍,7+1 套间

The shared trait is not "lacking a home" — it is using a constrained, impermanent space, yet still choosing to live intentionally inside it.

处在过渡性居住状态,但仍想认真生活的人。

1.3  Product Scope

产品范围

v0.3 focuses on living spaces, not workspaces. The reasoning: work spaces primarily serve organisational goals; living spaces primarily serve personal life. When users leave their professional identity behind, life actually unfolds — cooking, music, reading, rest, exercise, solitude, socialising. NestAI addresses lifestyle, not productivity.

Workstation / desk / creative-corner sub-scenes may be introduced in future, but the v0.3 product core is the residential space — the most complete container for lifestyle.

1.4  Pain Points

核心痛点
Pain 1
Users don't always know how they want to live. Many haven't been consistently asked: what do you actually want here? What should happen in this space? This can't be resolved by a one-time questionnaire — it must emerge through sustained dialogue and spatial decisions. 用户未必知道自己真正想怎么生活
Pain 2
Rich external references, no path to personal space. Users have seen countless beautiful rooms but can't judge: does this fit my dimensions? My budget? My rental rules? My body? They lack a translation layer. 有大量参考,但不知道如何落到自己空间里
Pain 3
Heavy constraints make standard design advice unexecutable. Cannot drill walls. Cannot change flooring. Limited budget. Uncertain tenure. Overlapping functions. Users need low-cost, low-risk, reversible, iterative spatial interventions — not renovation blueprints. 居住空间强约束,普通设计建议经常无法执行
Pain 4
No system that continuously learns the user. Generic LLMs answer one-off questions. They cannot reliably retain long-term lifestyle changes, per-intervention feedback, recurring spatial needs, or patterns of what repeatedly works — or fails — for a given person. 用户缺少一个持续学习自己的系统

1.5  Product Philosophy

产品哲学

Platform inspiration, design templates, branded content, and network aesthetics are not the antagonist. They are the system's nutrient — supplying style samples, spatial cases, object references, and lifestyle vocabulary.

But nutrient does not automatically grow into a lifestyle that belongs to a specific person. A genuine lifestyle must grow from a particular soil — the user's behaviours, personality, body, budget, spatial constraints, lighting, circulation, noise, privacy. Constraints are like the size of the pot: they limit what can grow, but also determine what will thrive here.

NestAI doesn't copy someone else's template into your room. It lets your lifestyle grow from your real space.

NestAI 不是把别人的模板搬进你的房间,而是让你的生活方式从你的真实空间里长出来。

1.6  Core Proposition

当前产品命题

NestAI sees your Nest, understands your lifestyle, then gives you your Next.

More precisely: through understanding the user's real space, life intentions, external inspiration, and personal constraints, NestAI translates a desired life into the next executable spatial intervention.

It doesn't only answer "who are you?" — it answers "given who you are now, the space you have now, and the life you want, what is the single most worthwhile thing to change next?"

02

Core Mechanism

核心产品机制

2.1  Lifestyle Growth System

生活方式生长系统

The core mechanism of NestAI v0.3 is the Lifestyle Growth System. This naming is more precise than "matching engine" — NestAI doesn't simply pair a user with a style or product. It helps the user externalise their lifestyle into physical space through repeated spatial interventions, choices, feedback, and adjustment.

The system operates in three layers:

Layer 1  ·  Understand the Soil

Real space, behaviours, personality, body, budget, tenure, constraints

理解土壤:真实空间与居住约束

Layer 2  ·  Absorb the Nutrient

Platform inspiration, design cases, spatial psychology, ergonomics, behavioural science, lifestyle knowledge

吸收养料:平台灵感与知识库

Layer 3  ·  Generate Growth Actions

Output the next executable spatial intervention — turning abstract lifestyle aspirations into real physical actions

生成生长动作:可执行的空间干预,让 lifestyle 从愿望变成行动

Internally, the system decomposes into four sub-mechanisms:

Life-Space Fit Detect match / mismatch between desired lifestyle and current space
Next Best Action Determine the single most valuable spatial action for the user right now
Intervention Plan Generate an executable spatial intervention plan at the right effort level
Living Memory Record interventions, choices, feedback, and changes — continuously updating the user's spatial profile

NestAI is a spatial agent that helps users grow their own lifestyle.

NestAI 是一个帮助用户生长出自己生活方式的空间 Agent。

2.2  Nest & Next

栖巢与下一步

Nest

The space the user currently occupies, owns in a limited sense, and is constrained by. Space type, dimensions, layout, existing objects, light, circulation, storage, noise, privacy — and the user's experience of all of it.

用户当下真实拥有、真实占据、真实受限的空间

Next

The next action derived from the current Nest and the user's life intention. Not a full renovation plan — the Next Best Spatial Action: the most worthwhile thing to change right now.

基于当前 Nest 和生活意图推导出的下一步行动

Next actions include six verbs:

MoveRelocate an object to change the spatial dynamic
RemoveEliminate a source of friction or visual noise
AddIntroduce a low-cost soft-furnishing or functional object
ReplaceSwap out something misaligned with the target lifestyle
ReframeRedefine the purpose of an existing zone
RitualizeEstablish a space-supported behavioural ritual

2.3  Lifestyle Components

Lifestyle 的构成

NestAI must avoid collapsing lifestyle into a style label. But it cannot assume users already have a clear, stable, nameable Routine. Most users don't say "I have a night-study routine." They say: I like gaming. I want to work out. I like quiet reading. I want my room to feel more like me.

These are not Routines — they are looser, more honest Habits, Interests, Tendencies, and Desires. So NestAI models lifestyle from low-threshold raw material:

Lifestyle  =  Habit + Interest + Preference + Body Feeling + Constraint + Aspiration + Routine

Routine 不是 lifestyle 的起点,而是 lifestyle 被空间持续支持之后逐渐长出来的结果。

Crucially: Routine is not the starting point of lifestyle — it is the outcome that emerges when Habits, Interests, and Aspirations are consistently supported by a space over time. NestAI doesn't assume a mature lifestyle; it helps grow one from loose, honest raw material.

03

Feature Design

功能设计

3.1  Memory & User Profile

记忆系统与用户画像

Memory is what distinguishes NestAI from a one-off AI design tool. It doesn't record isolated tags or objects. It records:

How users currently live, how they aspire to live, how their space changes, and whether those changes actually made a difference.

Five classes of information are tracked:

Current Lifestyle

Habits, Interests, Preferences, Body Feelings, and discovered Constraints

当前 Lifestyle:习惯、兴趣、偏好、身体感受、约束

Aspirational Lifestyle

What more / less they want to do, what new habits to build, how they want the space to feel, who they want to become

向往的 Lifestyle:想要成为怎样生活的人

Space Profile

Space type, objects, multi-use zones, light, circulation, fixed vs. movable elements, discovered constraints

空间画像:物理空间的动态理解

Intervention History

Every Move, Remove, Add, Replace, Reframe, Ritualize action the user has attempted

空间干预 Action 历史:每次空间行动的记录

Post-intervention Feedback

Did the user execute? How hard was it? Did it meet expectations? What worked, what didn't? Did new constraints emerge? Any regrets?

Action 结果与使用后反馈:干预是否真的发生了效果

Key principles: Memory is not a tagging system. It is not a recording burden. It comes from action. It simultaneously tracks the present and the aspiration. It serves the next intervention.

3.2  Space Profile

空间画像

The space profile does not try to fully model the room once upfront. It proceeds in three layers:

Physical fact recognition

Identify basic objects: bed, desk, chair, wardrobe, shelves, lamps, outlet positions, windows, doors, open floor area, usable wall zones, movable objects.

第一层:物理事实识别

Multi-function zone understanding

Real spaces — especially dorms and micro-apartments — are highly composite. The bed is also a game station, cinema, office, and social space. The desk is also a makeup table, dining area, and storage pile. NestAI records which behaviours co-inhabit each zone and where they conflict.

第二层:复合功能理解,不做过强功能划分

Constraints revealed through feedback

Constraints are not a cold-start form. They emerge as interventions are proposed and users respond: "can't drill," "can't afford," "can't move it alone," "dorm rules." The system writes these into the profile progressively.

第三层:从方案反馈中发现约束

The core question is not "what's in the room?" — it's "how do these objects, zones, and limits collectively shape the life the user wants to live?"

3.3  Knowledge Base

知识库

The knowledge base is the nutrient of the Lifestyle Growth System. Two types:

Design Knowledge

How to change a space — spatial design principles, small-space / dorm renovation methods, ergonomics, furniture scale, circulation, lighting, storage, soft-furnishing, colour, style, reversible low-cost interventions, platform templates and cases

设计类:空间可以怎么改?

Human Knowledge

Why changing a space changes a life — spatial psychology, environmental behaviour, behavioural science, habit formation, sleep / focus / exercise / social / recovery scene knowledge, sensory perception, social relations and dwelling

人的知识库:为什么这样改会影响人的生活?

Design knowledge alone produces renovation advice. Human knowledge alone produces lifestyle inspiration content. Both together produce interventions that are executable and connected to a real lifestyle.

3.4  Spatial Intervention Plan

空间干预方案

The intervention plan is the core feature of v0.3 and the highest-priority output in the Studio demo. It is not a detailed renovation report. It first shows the user a transformed life result:

If you do this — here is what your space and your life could look like.

Output formats, in order of visual priority for Studio demo:

Rotatable 3D isometric / panoramic preview

可旋转 3D 轴测模型 / 全景空间预览

Human-eye perspective renders / mood renders

人视角效果图 / 氛围渲染图

Before / after comparison

改造前后对比图

Floor plan, object list, action steps

平面图 / 物品清单 / 行动步骤

Plans come in three tiers:

Tier 01

Zero Cost

0 元调整

No purchasing — move, clear, reorganise, redefine usage rules

Tier 02

Low-cost Soft Furnishing

低成本软装

Lighting, fabric, plants, posters, small furniture, bedding — affordable, portable, reversible

Tier 03

Advanced Intervention

进阶改造

Fuller local arrangement — renders, shopping list, object combinations, layout adjustments, stronger style expression

Every plan includes one extremely light starter action — the first physical thing to do, as simple as standing up from a chair:

Example"First, clear the top-right corner of your desk."  /  "First, move the charger cable from the bed to the desk."  /  "First, turn off the overhead light and keep only your desk lamp."

Get the user moving first — and let the lifestyle grow from there.

先让用户动起来,再让生活方式长出来。

Each plan must state which Habit / Interest / Desire / Aspiration it serves — to ensure every spatial change is explicitly connected to a life goal, not just aesthetics. And shopping lists, when included, appear only after a user commits to a plan, and must explain why each object is needed, whether a zero-cost substitute exists, and whether it can travel to the next home.

04

User Journey

用户旅程

Space & Lifestyle Input

User photographs / scans / describes their space. AI conversation uncovers daily habits, interests, preferences, body feelings, and aspirations — without asking "what is your Routine?" but instead: "what do you most often do here? What would you like this room to support more? What moment in this space feels worst?"

4.1 数据录入:空间 & Lifestyle

Match Diagnosis

System diagnoses: where does the current space support the user's lifestyle? Where does it obstruct it? What is the most worthwhile problem to address first? Why?

4.2 匹配诊断

Spatial Intervention Output

The most critical value-delivery moment in the journey. The user sees a desirable spatial result — and understands the single, light action they can start with right now.

4.3 输出空间干预方案

User Action in Physical Space

The user selects a plan or a starter action and executes it in the real world. This is the defining difference from AI content generators — the goal is to get the user off the app and moving in their room.

4.4 用户行动

Intervention Follow-up & Analysis

Did you do it? Where was it hardest? Did it bring you closer to your target lifestyle? What couldn't you execute? What surprised you? What next? — Responses update Memory, refine profiles, and generate the next intervention.

4.5 干预回访与分析

Community Sharing & Peer Matching (Optional)

MVP scope: share before/after comparisons, lifestyle cards, "My Next" cards, object lists. Cautious about: follows, likes, public feeds, rankings. Potential peer matching: by similar habit, similar space constraint, similar life theme.

4.6 社区分享与同好匹配(Optional)
05

MVP Scope

MVP 边界

5.1  Studio MVP — Minimum Closed Loop

最小闭环

The Studio MVP must demonstrate a complete Lifestyle Growth System loop:

ScanUser photographs / scans their space via mobile. System obtains room photo, video, or panoramic material.
TalkAI dialogue understands current habits, interests, preferences, body feelings, and aspirational lifestyle.
DiagnoseSystem calls user data, space profile, and knowledge base to assess match / mismatch.
Output3D model / panorama + mood render + three-tier plan + action steps + lifestyle goal linkage + expandable shopping list.
ActUser selects a light starter action and executes it in physical space.
ReflectUser feeds back difficulty, result, expectation delta, blockers.
UpdateSystem forms change assessment, updates Memory, generates the next Next.

NestAI doesn't just generate plans in a virtual world — it uses virtual preview and action guidance to drive real change in a physical space.

5.2  Product MVP

产品 MVP

The product MVP proves: can NestAI translate a user's life intention into real change in their physical space?

Not currently prioritised: long-term spatial personality precision, community growth, cross-dwelling migration, full interior design generation capability.

5.3  Technical MVP

技术 MVP

Mobile-first vertical Web App. The camera is central to spatial input, and the interventions happen in the physical space — mobile is the right context to scan, receive a plan, act, and follow up.

Core stack

Mobile-first Web App · LLM API · Vision / image understanding · Mock user + space data · Iteratively expanding knowledge base

Minimum viable output

At least one visual output: render / isometric / panorama / before-after comparison. Minimum memory: current lifestyle, aspirational lifestyle, space state, intervention actions, user feedback.

Technical focus: run the full loop — Scan → Understand → Call knowledge base → Generate intervention → User feedback → Update Memory — not build all modules at once.

06

Open Questions

暂缓决策项

The following require further discussion between Vendow and Wenhuan. Not finalised in v0.3.

Memory Data Structure

Should the next version formally define data structures: UserProfile, SpaceProfile, LifestyleProfile, InterventionAction, Feedback, KnowledgeChunk? Core tension: how does Memory support long-term growth without increasing user recording burden?

Memory 数据结构

Intervention Plan Output Template

What exactly appears in the first screen? How are the 3D model, mood render, panorama, and three-tier plan organised? How does the light starter action connect to expanded detail? How does the user go from "reading the plan" to "acting now"?

空间干预方案输出模板

First Mobile Interaction Flow

Define five core screens: Welcome / Onboarding → Scan Space → Lifestyle Chat → Intervention Result → Feedback & Memory. Key challenge: keep the interaction light enough that users don't feel they're filling in forms or doing homework.

第一版移动端交互页面
07

v0.3 Definition

一句话定义

NestAI is a spatial agent that helps young renters and dormitory residents grow their own lifestyle. By understanding the user's real space, current and aspirational lifestyle, external inspiration, and personal constraints, it generates visualisable, executable, and feedbackable spatial intervention plans — driving users to externalise the life they want, step by step, into the real physical world.

NestAI 是一个帮助泛年轻租住人群与宿舍人群生长出自己生活方式的空间 Agent。它通过理解用户的真实空间、当前与向往的 lifestyle、外部灵感与个人约束,生成可视化、可执行、可反馈的空间干预方案,推动用户把想过的生活一步步外化到真实物理世界中。