Features

The technical details.

For those who want to know exactly how LifeTrace captures, stores, measures, and protects the traces of your life.

Data sources

What you can connect

Source	Method	Status
Gmail	OAuth2, incremental sync	Available
Google Calendar	OAuth2, event + location sync	Available
iMessage	macOS chat.db (Full Disk Access)	Available
Apple Photos	Photos.sqlite EXIF GPS clustering	Available
Google Timeline	On-device export + reverse geocoding	Available
Local files	Directory scan	Available
WhatsApp / Telegram / Signal	Chat export import	Planned
Outlook	Microsoft Graph API	Planned
Voice notes	Whisper transcription	Planned

Journal

A narrative of your life, told back to you

Every day, LifeTrace writes a short reflection of who you were that day — grounded in what your traces actually show, not what an AI imagines about you.

Daily entries — 2–3 sentences anchored by GPS, calendar events, and conversations

Weekly + monthly digests — higher-altitude reflections built from the daily layer

Threads view — recurring themes traced across days, with rank changes over time

History grid — every day in your corpus shown at a glance, GitHub-style heatmap

Backfill any range — generate years of historical entries in one pass

Search + tag filter — find the day you were sick, the week of a wedding, the month you started something new

Sample · Tuesday, April 21

This was a wedding day folded around real life: in the morning Gaby got married at the courthouse, and the rest of the day was split between childcare with Angelica, quick personal appointments, and the background pull of work messages that never fully stopped.

weddingfamilylogisticscelebration

Places

Where you've been, drawn as a trace

Combines GPS history from Google Timeline and geotagged Apple Photos into a single map. Markers cluster intelligently. Faint chronological lines show the path between visits. Click any spot to read the journal entry from that day.

Reverse geocoding

Coordinates resolved to neighborhood + city names so the LLM grounds entries in real geography.

Photo clustering

Geotagged photos collapse into place visits by 30-min / 200-m proximity — one trace per outing, not per shutter click.

Tag filtering

Personal vs. work vs. family — see only the geography of one part of your life.

Corpus health

Six dimensions

LifeTrace continuously scores your corpus to show how complete the picture is — and what's missing.

Volume — How much self-authored content you have

Modality — Diversity of source types

Dialogue — Conversational vs. monologue content

Temporal — Coverage across time, gap detection

Breadth — Range of topics and interests

Emotion — Sentiment variance — a real person isn't always neutral

Sample

Volume

Modality

Dialogue

Temporal

Breadth

Emotion

Desktop app

Native macOS, Linux, Windows

Packaged as an Electron app: drag, double-click to launch, native menus. Full Disk Access granted to LifeTrace.app — never to your terminal.

AI providers

Bring your own LLM

Ollama locally for free. OpenAI, Anthropic, or Google via API key. Codex subscription via OAuth. Any OpenAI-compatible endpoint — including your own private models.

Privacy

Your machine, your rules

Self-hosted on your hardware. Local SQLite storage. Consent receipts on every item. PII detection for third-party content. Open source — inspect everything.

Waitlist

Start building your corpus.

Join the waitlist for early access.