docs

Documentation

v1.3.0last modified 2026-03-29
01

What is Ekachit?

Ekachit is a focus intelligence platform that connects your laptop, phone, and calendar into a single intelligence layer. It passively observes how you spend time, identifies patterns, and coaches you toward deeper work.

// what Ekachit does
1. Tracks focus sessions — desktop + phone + calendar
2. Builds your focus profile — 6D spider chart (ML-powered)
3. Sends smart nudges — 14 types of contextual coaching
4. Protects deep work — phone guardian + meeting cost analysis
5. Guides daily rituals — morning plan + evening review
ℹ Ekachit never records or transmits window titles, URLs, keystrokes, screenshots, or meeting content. An optional local-only feature can categorize window titles on your device — these titles never leave your machine.
02

System Architecture

Ekachit collects anonymous activity metadata from your devices, processes it through our privacy-first pipeline, and delivers personalized focus insights.

Mobile App — iOS + Android. Focus timer, tasks, nudges, insights, daily rituals.
Desktop Agent — macOS menu bar. Silent activity categorization.
Integrations — Google Calendar, Slack, Jira. Meeting sync, status, tasks.
03

5-Minute Setup

Get your first focus insights in 5 minutes. Here's the full setup path:

01
Create Account~30s

Sign up at ekachit.app/signup. Email + password or use Google/GitHub/Apple SSO.

→ detailed guide
02
Install Mobile App~1m

Download from App Store (iOS) or Play Store (Android). Sign in with your account.

03
Connect Google Calendar~1m

Settings → Integrations → Google Calendar → Authorize. Read-only access — we only see meeting times, never content.

→ detailed guide
04
Install Desktop Agent~2m

Download the macOS agent. Grant Accessibility API permission. It runs silently in your menu bar.

→ detailed guide
05
Start Your First Session~30s

Tap 'Start Focus' in the app. Put your phone down. Work. Ekachit handles the rest.

✓That's it. After your first focus session completes, you'll see your initial insights within minutes. The ML engine needs ~3 days of data to generate your full 6D focus profile.
04

Create Your Account

You can create an account in three ways:

→ Email + Password — classic, works everywhere
→ Google SSO — one-click, auto-links calendar
→ GitHub SSO — popular with developers
→ Apple SSO — private relay email supported
💡 If you sign up with Google, calendar integration is automatically connected — one less step.
account requirements
email:valid, unique
password:8+ chars, 1 uppercase, 1 number
min_age:13 years
free_accounts:1 per person
05

Plans & What You Get

FREE$0 /forever
✓ Focus timer + phone detection
✓ Google Calendar sync
✓ Up to 5 active tasks
✓ Basic daily insights
✓ Morning plan (card mode)
PRO$4.99/mo
✓ Everything in Free
✓ Unlimited tasks + templates
✓ Desktop agent + auto-tracking
✓ ML coaching + 6D spider chart
✓ Slack + Jira integration
✓ Guided morning ritual + evening review
✓ Priority support

# Desktop agent, Slack, and Jira require Pro plan.
# Cancel anytime. Access continues until end of billing period.

06

Desktop Agent Setup

The macOS desktop agent silently categorizes your active application into focus categories. Full setup guide:

⌘
Desktop Agent — macOS Setup Guide
Download, install, permissions, configuration, troubleshooting
→PRO PLAN REQUIRED
07

Phone App Setup

Install the mobile app and configure phone-aware focus tracking.

📱
Phone Setup — iOS & Android
Installation, permissions, what's tracked, Focus Guardian
→
08

Google Calendar Integration

Sync your calendar to enable meeting cost analysis and focus window detection.

📅
Google Calendar — Setup Guide
OAuth, scopes, what syncs, meeting cost calculation
→
09

Slack Integration

Connect Slack for automatic DND sync and focus status updates.

#
Slack — Setup Guide
Bot install, DND sync, focus status, notification batching
→PRO PLAN REQUIRED
10

Jira Integration

Import tasks from Jira and map estimation data for ML coaching.

◆
Jira — Setup Guide
API token, project sync, task import, estimation mapping
→PRO PLAN REQUIRED
11

Focus Sessions

Focus sessions are the core unit of Ekachit. When you start a session, the system begins tracking across all connected devices.

session object
start_time:ISO 8601 timestamp
duration:elapsed time
linked_task:optional task reference
phone_pickups:count during session
recovery_cost:estimated minutes lost
focus_score:0-100 (ML-computed)
category_breakdown:IDE: 65%, browser: 20%, etc.

During a session, Ekachit monitors:

desktop → which category of app is active (IDE, browser, meeting, etc.)
phone ──→ detects when you open the Ekachit app during a focus session (does not monitor other apps, screen time, or general phone usage)
calendar → whether a meeting is scheduled during the session
⚠ Each phone pickup during focus costs an average of 23 minutes of recovery time. Ekachit tracks this and shows cumulative cost.
12

Task System

Tasks are what you're working on during focus sessions. Ekachit uses your task history to predict future durations.

task schema
title:string (required)
time_estimate:optional, ML-assisted
actual_time:auto-computed from sessions
status:todo | in_progress | done
subtasks:up to 30 checkable items
confidence:ML prediction confidence (0-1)
source:manual, Jira import, or template

After ~10 completed tasks, the ML engine begins generating estimation predictions. At ~30 tasks, it reaches 94% accuracy.

13

Nudge System

Ekachit sends 14 types of contextual nudges based on your behavior patterns:

focus: session_reminder, break_suggestion, phone_pickup_alert, streak_celebration
meetings: true_cost_warning, prep_time_nudge, back_to_back_alert
tasks: stale_task_reminder, estimation_coach, deadline_risk
habits: morning_plan_prompt, evening_review_prompt, weekly_summary
comms: batch_notification_digest

Nudges are delivered via push notification and in-app overlay. You can disable any category individually in Settings → Notifications.

14

Spider Chart & 6D Profile

The 6D focus profile measures your work patterns across six dimensions, updated daily by the ML engine:

dimensions
deep_work:% of time in uninterrupted focus (>25 min blocks)
quality:focus score consistency across sessions
efficiency:actual vs estimated task completion ratio
focus:inverse of context-switching frequency
balance:work/break ratio vs healthy baseline
discipline:adherence to planned schedule + phone-free streaks

The spider chart requires ~3 days of tracking data before the first profile is generated. Accuracy improves continuously as more data is collected.

# Spider chart is a Pro feature.

15

Data Flow & Encryption

All data flows through encrypted channels with minimal server-side exposure:

security measures
transport:Industry-standard encrypted connections
at_rest:Encrypted storage with per-user data isolation
auth_tokens:OS-level secure storage on all platforms
config:Owner-only file permissions
api_keys:Stored server-side, never on client
ml_input:Anonymized aggregate metrics only

Per-user data isolation ensures every query is scoped to the authenticated user. No user can access another user's data.

16

On-Device Processing

The most sensitive computation — cross-device correlation — happens entirely on your device.

// what runs on-device
→ Phone pickup ↔ desktop activity correlation
→ Focus session boundary detection
→ Real-time phone guardian alerts
// what runs server-side
→ ML model inference (anonymized aggregates)
→ Push notification delivery
→ Calendar sync (OAuth token refresh)

Raw behavioral data stays on your device. Only anonymized aggregate metrics are sent to our servers for focus scoring.