AI-Powered Task Creation: From Natural Language and Images to Structured Tasks

Task capture shouldn't require typing every detail manually. When you photograph a whiteboard full of meeting notes, your task manager should extract the action items automatically. When you describe your day's plan ("Book Friday's flight and pack bags for the weekend"), it should create separate structured tasks instantly.

AI-powered task generation makes this possible. Instead of:

1. Read whiteboard photo
2. Type first task name
3. Add tags manually
4. Set dates manually
5. Add properties
6. Repeat 10 times...

You:

1. Take photo (or describe in natural language)
2. Review AI-generated tasks
3. Create all at once

To-Track's AI task generator (Pro feature) transforms text and images into structured tasks with intelligent suggestions for names, tags, dates, and properties. It's not magic—it's practical AI that eliminates the busywork of manual task entry.

This guide shows you how to use AI task generation for brainstorming, meeting notes, visual capture, and rapid task creation.

The Two Input Methods

To-Track's AI supports two types of input:

1. Natural Language (Text Prompts)

Describe what you need to do in plain English:

Example prompts:

• "Book flight for Friday and pack bags for weekend trip"
• "Schedule dentist appointment, pick up prescriptions, and call insurance about claim"
• "Write blog post about AI features, create social media graphics, and schedule promotion"
• "Buy groceries: milk, eggs, bread, chicken. Also pick up dry cleaning"

AI output: Multiple tasks with intelligent parsing:

• Separates distinct actions
• Infers dates from context ("Friday" → this Friday's date)
• Suggests relevant tags ("travel" for flights, "errands" for groceries)
• Recommends properties if applicable

2. Vision (Image Analysis)

Upload up to 3 images containing task information:

Supported image types:

• Whiteboard photos (meeting notes, brainstorming sessions)
• Screenshots of lists (emails, documents, Slack messages)
• Handwritten notes
• Physical sticky notes
• To-do lists from other apps

AI output: Extracts text, identifies tasks, and structures them with tags/properties.

How AI Task Generation Works

The Workflow

1. Input: Tap "Create with AI" in Add Task popup
2. Choose method:
   • Text: Type natural language description
   • Images: Select up to 3 photos from gallery or camera
3. AI processing: Cloud-based analysis (encrypted, ephemeral)
4. Review: Preview screen shows generated tasks
5. Edit: Modify, delete, or add tasks as needed
6. Create: Tap "Create All Tasks" to add to your database

Context-Aware Intelligence

To-Track's AI isn't generic—it knows your task system:

Available tags: AI sees your existing tags and suggests them for new tasks

• You have tags: "work", "personal", "urgent"
• Prompt: "Call client about proposal"
• AI suggests: Tag = "work"

Available properties: AI recommends properties that match task type

• You've tracked "Duration", "Difficulty", "Mood" for workouts
• Prompt: "Morning workout"
• AI suggests: Properties = Duration, Difficulty

Existing tasks: AI avoids creating duplicates

• You already have task "Book flight"
• Prompt: "Book flight and rent car"
• AI creates only "Rent car" (suggests updating existing "Book flight" instead)

Smart Date Parsing

AI understands temporal context:

Relative dates:

• "Friday" → This Friday (if today is Monday-Thursday) or next Friday (if today is Friday-Sunday)
• "Next week" → Monday of next week
• "Tomorrow" → Actual tomorrow date
• "This weekend" → Saturday of current week

Specific dates:

• "March 15" → March 15, 2026 (current year assumed)
• "Jan 1st" → January 1 of next occurrence

Time of day:

• "Meeting at 3 PM" → Sets start time to 3:00 PM
• "8 AM workout" → Sets start time to 8:00 AM

Use Case 1: Meeting Notes (Vision)

Scenario: You ran a brainstorming meeting, filled a whiteboard with ideas, and need to convert them to tasks.

Whiteboard content (handwritten):

Launch new product
- Design landing page
- Write product copy
- Set up email sequence
- Create demo video
- Schedule social posts
- Q2 deadline!

Follow-ups:
- Email John about pricing
- Review competitor sites
- Book designer for mockups

Process:

1. Take photo of whiteboard
2. Open To-Track → "Add Task" → "Create with AI"
3. Upload whiteboard photo
4. AI processing...

AI-generated tasks:

1. "Design landing page"
   - Tags: work, design
   - End time: Q2 (June 30, 2026)
   - Parent: "Launch new product"

2. "Write product copy"
   - Tags: work, content
   - End time: Q2
   - Parent: "Launch new product"

3. "Set up email sequence"
   - Tags: work, marketing
   - End time: Q2
   - Parent: "Launch new product"

4. "Create demo video"
   - Tags: work, video
   - End time: Q2
   - Parent: "Launch new product"

5. "Schedule social posts"
   - Tags: work, social
   - End time: Q2
   - Parent: "Launch new product"

6. "Email John about pricing"
   - Tags: work
   - No date (follow-up ASAP)

7. "Review competitor sites"
   - Tags: work, research

8. "Book designer for mockups"
   - Tags: work, design

Review screen: Shows all 8 tasks with suggested tags, dates, and parent task relationship.

Actions:

• Tap any task to edit details
• Delete tasks that aren't needed
• Add additional tasks manually
• Tap "Create All Tasks" to add to database

Result: 30 seconds of review instead of 10 minutes of manual typing.

Use Case 2: Natural Language Batch Creation

Scenario: Monday morning planning—you need to capture the day's tasks quickly.

Prompt:

"Schedule team standup at 9 AM, review pull requests before lunch, finish blog post about AI features this afternoon, call dentist to book appointment for next week, and pick up dry cleaning after work"

AI processing: Parses five distinct tasks from the run-on sentence.

AI-generated tasks:

1. "Team standup"
   - Tags: work, meeting
   - Start time: Today 9:00 AM
   - End time: Today 9:30 AM (inferred 30 min duration)

2. "Review pull requests"
   - Tags: work, code
   - End time: Today 12:00 PM (before lunch)

3. "Finish blog post about AI features"
   - Tags: work, content
   - Start time: Today 1:00 PM (afternoon)

4. "Call dentist to book appointment"
   - Tags: personal, health
   - No time (ASAP, flexible)

5. "Pick up dry cleaning"
   - Tags: personal, errands
   - End time: Today 6:00 PM (after work, assumed 5 PM end of workday)

Intelligence demonstrated:

• Separated 5 tasks from 1 sentence
• Inferred meeting duration (standup = 30 min)
• Understood "before lunch" = end time 12 PM
• Recognized "afternoon" = start time 1 PM
• Distinguished work vs. personal tasks via context

Use Case 3: Email/Screenshot Task Extraction

Scenario: Client sends email with 7 deliverables. You screenshot the email and upload to To-Track.

Email screenshot content:

Hi [Name],

For the website redesign project, we'll need:
1. Homepage mockup (due Feb 1)
2. About page design (due Feb 1)
3. Contact form implementation (due Feb 5)
4. Blog template design (due Feb 8)
5. Mobile responsive testing (due Feb 10)
6. Final review meeting (week of Feb 12)
7. Launch on Feb 15

Let me know if you have questions!

Process:

1. Screenshot email
2. To-Track → "Create with AI" → Upload screenshot
3. AI extracts list items

AI-generated tasks:

1. "Homepage mockup"
   - Tags: client-work, design
   - End time: Feb 1, 2026

2. "About page design"
   - Tags: client-work, design
   - End time: Feb 1, 2026

3. "Contact form implementation"
   - Tags: client-work, development
   - End time: Feb 5, 2026

4. "Blog template design"
   - Tags: client-work, design
   - End time: Feb 8, 2026

5. "Mobile responsive testing"
   - Tags: client-work, testing
   - End time: Feb 10, 2026

6. "Final review meeting"
   - Tags: client-work, meeting
   - Start time: Feb 12, 2026 (week of)

7. "Launch website"
   - Tags: client-work
   - Start time: Feb 15, 2026

Result: Client's 7-item list becomes 7 tasks in your system instantly, with dates pre-filled.

Use Case 4: Grocery/Shopping Lists

Scenario: You keep a running note of groceries needed and want to convert it to tasks.

Note content:

Groceries:
- Milk
- Eggs
- Bread
- Chicken breast
- Spinach
- Coffee

Also:
- Pick up package at post office
- Return library books

Prompt (typed or photo of note):

"Buy milk, eggs, bread, chicken breast, spinach, coffee. Pick up package at post office and return library books."

AI-generated tasks:

1. "Buy groceries"
   - Tags: errands, shopping
   - Notes: milk, eggs, bread, chicken breast, spinach, coffee

   (OR if you prefer separate tasks:)

1. "Buy milk"
   - Tags: errands, groceries
2. "Buy eggs"
   - Tags: errands, groceries
3. "Buy bread"
   - Tags: errands, groceries
... (AI can detect preference based on your existing tasks)

7. "Pick up package at post office"
   - Tags: errands

8. "Return library books"
   - Tags: errands

Preference detection: If your history shows consolidated grocery tasks, AI groups items. If you track individual items, AI creates separate tasks.

Use Case 5: Travel Planning

Scenario: Planning a trip with many moving parts.

Prompt:

"Planning trip to NYC March 10-15. Need to book flights, reserve hotel in Manhattan, buy Broadway tickets for Hamilton, make dinner reservation at The Modern, pack bags, arrange pet sitter, and set email auto-reply."

AI-generated tasks:

1. "Book flights to NYC"
   - Tags: travel, planning
   - Start time: March 10, 2026
   - End time: March 15, 2026
   - Notes: Round trip

2. "Reserve hotel in Manhattan"
   - Tags: travel, accommodation
   - Start time: March 10 (check-in)
   - End time: March 15 (check-out)

3. "Buy Broadway tickets for Hamilton"
   - Tags: travel, entertainment

4. "Make dinner reservation at The Modern"
   - Tags: travel, dining
   - Notes: During trip dates

5. "Pack bags for NYC trip"
   - Tags: travel
   - End time: March 9 (before departure)

6. "Arrange pet sitter"
   - Tags: personal
   - End time: March 9
   - Notes: March 10-15 dates

7. "Set email auto-reply"
   - Tags: work
   - End time: March 10 (before leaving)

Intelligence:

• Extracted date range (March 10-15)
• Applied dates logically (pack before trip, not during)
• Inferred task categories (travel vs. work vs. personal)

Advanced Features

Multi-Image Upload (Up to 3)

Combine multiple sources:

• Image 1: Whiteboard with project roadmap
• Image 2: Screenshot of email with deadlines
• Image 3: Photo of handwritten notes

AI processes all three and merges information intelligently.

Task Preview & Editing

Before creating tasks, the preview screen lets you:

• Edit task names: Clarify AI's interpretation
• Modify tags: Add/remove suggested tags
• Adjust dates: Correct or add start/end times
• Delete tasks: Remove tasks you don't need
• Add tasks: Manually add tasks AI missed

Example edit:

AI suggests: "Meeting with John"
You edit: "Quarterly review meeting with John (Q1 budget discussion)"

Batch Properties Assignment

If AI detects a task type you commonly track properties for:

• "Workout" → Suggests Duration, Reps, Weight properties
• "Reading session" → Suggests Quantity, Duration properties

You can accept or remove property suggestions before creating tasks.

Parent-Child Relationships

AI infers task hierarchies:

• Prompt: "Launch product: design landing page, write copy, create demo"
• AI creates:
  • Parent: "Launch product"
  • Children: "Design landing page", "Write copy", "Create demo"

All children automatically link to parent.

Privacy & Data Security

To-Track prioritizes privacy, even for AI features:

How It Works

1. Local first: Images/text prepared on device
2. Encrypted upload: Sent to Supabase Edge Function via HTTPS
3. Cloud processing: OpenRouter API (Claude Haiku model) analyzes input
4. Immediate deletion: All data deleted after response
5. No storage: Images and prompts are NOT retained

What's Sent

• Image data (base64 encoded)
• Text prompts
• Available tags/properties (for context)

What's NOT Sent

• Your full task database
• Task contents (only tag/property names for suggestions)
• Personal information beyond the prompt

Opting Out

AI task generation is opt-in (Pro feature). Standard To-Track remains 100% offline with zero AI processing.

Pricing: To Track Pro

AI task generation is part of To Track Pro subscription:

Includes:

• AI task creation (text + images)
• AI list filtering (natural language queries)
• AI statistics generation (custom charts)

Pricing: ~$4.99/month or $39.99/year (managed via RevenueCat)

Free tier: Core To-Track features (custom states, properties, time tracking, offline storage) remain free without AI.

Comparing AI Task Generation

How does To-Track stack up against competitors?

Feature	To-Track Pro	Notion AI	Todoist AI	ClickUp AI
Vision (image input)	✅ Up to 3 images	❌ No	❌ No	❌ No
Natural language	✅ Full sentences	✅ Limited	⚠️ Basic	✅ Yes
Batch creation	✅ Unlimited tasks	⚠️ Limited	⚠️ Limited	✅ Yes
Context awareness	✅ Tags, properties	⚠️ Basic	❌ No	⚠️ Limited
Smart date parsing	✅ "Friday", "next week"	⚠️ Limited	⚠️ Specific dates only	✅ Yes
Privacy	✅ Ephemeral processing	❌ Trains on data	❌ Cloud storage	❌ Cloud storage
Offline fallback	✅ Manual entry still works	❌ Requires internet	❌ Requires internet	❌ Requires internet

Verdict: To-Track's vision support and privacy-conscious design are unique advantages.

Tips for Best Results

1. Be Specific in Prompts

Vague: "Handle project stuff" Specific: "Review project timeline, email team updates, and schedule next sprint planning meeting for Thursday"

Specificity = better task separation and tagging.

2. Include Dates When Known

Without dates: "Book flight" With dates: "Book flight for Friday March 15 leaving morning"

AI uses date context for better scheduling.

3. Use Clear Images

For vision input:

• Good lighting (no glare or shadows)
• Legible handwriting or typed text
• Close-up shots (avoid distant whiteboard photos)
• One topic per image (don't mix unrelated lists)

4. Review Before Creating

AI is smart but not perfect:

• Check dates (especially "Friday" could be this week or next)
• Verify tags match your system
• Confirm task separation (sometimes AI merges tasks that should be separate, or vice versa)

5. Combine with Manual Entry for Complex Tasks

Use AI for bulk capture, then manually:

• Add detailed notes
• Set custom states beyond default
• Assign parent-child relationships AI missed

Conclusion: Task Capture at the Speed of Thought

Manual task entry is slow:

• Type task name
• Select tags
• Set dates
• Add properties
• Repeat...

AI task generation is instant:

• Describe or photograph
• Review suggestions
• Create all at once

For meeting notes, email action items, brainstorming sessions, and travel planning, AI eliminates the busywork. You focus on thinking about what needs to be done; AI handles the data entry.

To-Track's vision support (unique among task managers) makes whiteboards, sticky notes, and handwritten lists instantly digital. Combined with natural language processing and context-aware suggestions (tags, properties, dates), task capture becomes effortless.

Stop typing tasks one by one. Start generating them in batches with AI.

FAQ

Q: Does AI work offline? A: No—AI processing requires internet for cloud API calls. However, offline fallback to manual task entry always works.

Q: What image formats are supported? A: JPEG, PNG (standard photo formats). Max 3 images per request.

Q: Can AI handle handwritten notes? A: Yes, but legibility matters. Clear handwriting works well; cursive or messy writing may have lower accuracy.

Q: Does AI understand my existing tags/properties? A: Yes. AI sees your current tags and properties to suggest them for new tasks, ensuring consistency.

Q: What if AI creates duplicate tasks? A: The preview screen shows all tasks before creation. Delete duplicates before tapping "Create All Tasks."

Q: Can I edit AI-generated tasks after creation? A: Absolutely. AI-generated tasks are identical to manually created tasks—fully editable.

Q: What languages does AI support? A: Currently English. Future updates may expand language support.

Q: How many tasks can AI create at once? A: No hard limit, but practical maximum is ~20-30 tasks per request for review manageability.

Q: Is my data used to train AI models? A: No. To-Track uses OpenRouter with ephemeral processing—data is deleted immediately after response. Your tasks are NOT used for model training.

---

Ready to experience AI task creation? Upgrade to To Track Pro and transform photos and descriptions into organized tasks instantly.