AtlaSelf
Try AtlaSelf

Personal AI. Built around you.

Your personal AI consilium.

Specialist AI agents. Your data, reflections and goals. One connected response.

Try AtlaSelf
AtlaSelf Body Atlas illustration, with sleep and daytime signals connected to the human figure

01

One question.The right perspectives.

Your context brings the right AI perspectives together.

Your Atlas

  • Sleep Score
  • Night HRV
  • Resting HR
  • Steps
  • Recovery & sleep
  • Nutrition
  • Movement
  • Habits

One clear next step

Earlier wind-down

Finish work 30 minutes earlier on weekdays.

Your feedback shapes the next step.

Personalization begins before the answer.

  1. A growing personal context

    Your goals, limitations, reflections and experiment history form the starting point. Missing information stays visible.

  2. Data prepared for understanding

    Statistical preprocessing puts changes beside your own ranges, trends and data coverage.

  3. The relevant history

    Retrieval brings relevant observations and past outcomes into the current question, rather than treating every conversation as a fresh start.

  4. A prompt built for the task

    The question, prepared data, relevant context and each agent’s role are assembled into structured prompts.

  5. Specialists, then synthesis

    Routing selects a direct response or a consilium. The coordinator may request another specialist round before the final answer.

  6. Experience becomes context

    Your feedback and experiment outcomes can inform later conversations. Personalization comes from context; it does not mean training a separate model for every person.

02

A small experiment. Your next step.

Try a change. Notice what happens. Bring the experience back to your AI.

Earlier wind-down

Day 4 of 7

My observation

A calmer evening. Easier to switch off.

  1. AI
  2. Experiment
  3. Your feedback
  4. Next conversation
Next: signals and Daily Trace

03 / Beyond another dashboard

Signals worth your attention.

You already collect the numbers. AtlaSelf highlights meaningful changes against your own history, so you have somewhere useful to start.

Signals depend on the available history and the quality of the source data. A signal is an invitation to explore.

Night HRV: 32 ms, 27.3% below the previous 30-reading baseline; usual range 36–52 ms. The card explains overnight HRV and its recovery context.

04 / Daily Trace

Your days leave a trace.

A few minutes to reflect can give the numbers a story. The late meeting, the walk, the evening that felt different: keep the context you would otherwise forget.

A moment from the day

Finished work earlier and walked home. I still felt tired, but the evening was less rushed. I would like to try this again tomorrow.

Journal excerpt over two days: one expanded reflection with metric values and changes, and two collapsed notes with metric indicators.

Analytics / Four perspectives

Understand yourself
from four perspectives.

Explore your data, your days and your experiments.

01

Connected metrics

See which measurements move together over time.

Connected metrics

How it works
Compare 30 paired nights of sleep HR and night HRV, each scaled to its own average and variation.
What you can learn
In this real example, lower sleep HR tends to coincide with higher night HRV (Spearman ρ = −0.66). This observation does not establish a cause.

02

Easier & harder days

Compare the patterns that show up on different kinds of days.

Easier & harder days

How it works
The original mobile chart compares metric patterns on either side of a shared axis.
What you can learn
See the directions associated with each group of days. Segments show relative weights within this example; they do not establish a medical score or a cause.

03

Your experiments

Compare your check-ins, feedback and measurements.

Your experiments

How it works
Bring the recorded measures, completion marks and feedback from an experiment into one view.
What you can learn
Review what changed, how consistently you followed the plan and what you noticed yourself.

04

Signals & themes

Explore which journal themes coincide with changes in your signals.

Signals & themes

How it works
Synthetic example: 8 signal days; late dinner, 4 unfavorable co-occurrences; evening walk, 1 unfavorable and 3 favorable; work strain, 2 neutral decreases.
What you can learn
Compare direction and valence without treating every decrease as worse. Co-occurrence helps you investigate without proving a cause.

06 / A personal project

I turned 40. I started asking better questions.

I had years of wearable data, plenty of habits and one persistent question: how could I turn all of this into something useful for my wellbeing?

At first, I collected screenshots and explained my context to ChatGPT and Claude. Each conversation meant putting the picture together again.

I started building AtlaSelf so that the picture could grow over time: data, daily life, experiments and AI, connected around one person.

— The creator of AtlaSelf

A few things you may be wondering.

Can I use AtlaSelf today?

AtlaSelf is in development. You can join the early-access list and, if you wish, try beta versions as they become available. Joining the list does not unlock the app or guarantee a particular release date.

Do I need a watch?

A wearable can add useful history, but manual entries and daily reflections also provide context. What the AI can say depends on the data you have.

Where can my data come from?

The application currently supports Android Health Connect, manual entry and supported file imports, including FIT and Apple Health exports. This does not mean a direct Garmin or Apple account integration.

Does every answer involve every agent?

No. Routing depends on the question. The personal assistant can answer directly; specialist and consilium routing provide additional perspectives when available for the selected access tier.

Is this medical care?

AtlaSelf is a wellbeing project. It supports reflection and exploration of everyday habits; it does not provide diagnosis, prescriptions or a replacement for professional care.

What happens to my email?

Your address is stored for early-access contact, separately from application health data. It is not sent to AI. The privacy notice explains retention and how to request deletion.

Be part of the beginning

Your next chapter starts with context.

Join the early-access list. We will let you know when access opens.

Your email is only used to contact you about early access. While registration is closed, this form does not submit or store it.

Read the privacy notice