Transparency
How Witara thinks
No hidden algorithms. No language model guessing. A deterministic engine with visible rules, explainable scores, and reasons you can read.
We gather your context
Tasks, events, notes, finances, and reminders flow into a unified picture. Each item belongs to a life context (Work, Health, Finance, Personal, Learning, Home) that carries its own weight.
We score every item
Each task, event, and transaction receives a score based on: deadline proximity, user-defined priority, behavioral patterns (completion rate, postponement history), and context weight. No randomness. No black box.
We learn from your behavior
When you complete a task, your momentum in that context rises. When you postpone, the system notes the pattern. Over time, Witara learns your rhythms: when you focus best, which areas get attention, and where things drift.
We surface what matters
From all scored items, Witara selects one priority, two to five recommendations, and one insight. Focus mode narrows this to just the priority. Do-not-disturb suppresses entire contexts on your quiet days.
We explain every decision
Each recommendation carries reasons: "Due today," "This context is high priority," "Tasks here are often delayed." You can always ask why something was surfaced and correct the system if it got something wrong.
Score modifiers
These are the exact factors that influence how items are ranked. Nothing hidden.
Deadline proximity
Overdue: +6, Today: +4, Tomorrow: +3, This week: +1
Context weight
Primary context: 2.0x, Watched: 1.5x, Other: 1.0x
Performance
High completion rate boosts score; frequent postponement lowers it
Momentum
Recent activity in a context temporarily amplifies its weight
Weekly focus
Focused context gets a 2.2x weight boost for the week
Repeated postpone
-1 penalty when an item has been postponed 2+ times
We use math, not magic. You can see exactly why.