I Built The Laziest Task Management System… And It Actually Works

I hate project management tools. There, I said it. I also obsess over them. Productivity has been a lifelong fixation for me, probably because I’m naturally not that productive. So I’ve always been hunting for that perfect system to fix me.

For years I’ve tried them all. Asana, Trello, ClickUp, Monday, TickTick, Google Tasks, Apple Notes, Notion, and plain paper notebooks. With Notion, I went completely overboard. I learned the entire platform just to build a “second brain” productivity system from scratch. Every time the pattern repeats: initial excitement, followed by about three weeks of diligent updates, then the slow drift back to my chaotic ways.

The tools weren’t working against me. The problem was that I would become obsessed with the system itself. I wanted the perfect tags, the perfect Kanban board, the perfect methodology. I convinced myself that if I just built the perfect system, then the tasks would magically get done. So I’d procrastinate on actual work by obsessing over the system. Building the system became my task, which was the perfect way to avoid my real tasks.

The problem wasn’t the tools themselves. The problem was me.

My brain refuses to work in neat little boxes. I think in bursts, make connections randomly, and often need to talk through my process before I understand what I’m actually trying to accomplish. Traditional PM tools demand structure before insight, and that’s backward for how I operate.

I almost gave up on the idea of having a system at all. I operated from memory and my inbox for a while, because that way I couldn’t procrastinate tasks in favor of system-building if I didn’t have a system. But that approach was far from perfect. Things constantly fell through the cracks.

The breakthrough came unexpectedly. I had built a health management system with Claude that was working surprisingly well. For the first time ever, I was compliant with tracking habits and daily journaling. This made me wonder: “Why am I so consistent with this system when I’ve failed with every other one? Where else could I apply these principles?”

Then it dawned on me. In the past, when I got overwhelmed, I would sometimes open up a chat and just word-vomit everything in my mind, asking the AI to help organize my thoughts and tasks. I’d never thought of turning that occasional coping mechanism into an actual system. But seeing my success with the health system, I decided to tackle the holy grail. Project management. The system I’d never been able to conquer.

What if I intentionally built a project management system inside my AI conversations? One that adapted to my thinking style instead of forcing me to adapt to it?

How It Works

My system is embarrassingly simple, which is probably why it works for me where complex tools failed. The entire structure consists of just three text files:

1. Active Tasks Document: Everything I need to do, organized and prioritized
2. Completed Tasks Document: An archive of everything I’ve finished
3. Custom Instructions: Guidelines for how Claude should handle my task data

That’s it. No complex boards, no multiple categories, no tagging systems to maintain, no linked databases, no integrations. Just three text files.

The real magic? I talk to it. I don’t type tasks into fields or drag cards between columns. I record voice notes, sometimes clear and specific, often rambling streams of consciousness, and Claude turns them into organized tasks.

“Here’s what I’m thinking about the AI implementation for that finance client… I need to help them find the ideal tool for their reporting needs. Let me work through the transcripts from our discovery call. I should probably also build them a prompt library for their specific use cases. And I need to schedule a follow-up call to review their team’s initial experiences…”

Claude extracts the actual tasks from my verbal chaos and adds them to the Active Tasks document without me having to structure anything.

The system accepts virtually any input format. If I take handwritten notes in a meeting, I just snap a photo and send it to Claude. Someone sends me a task via WhatsApp? Screenshot, send to Claude. I see something interesting on the street that triggers a project idea? Take a picture, feed it to the system. This multimodal capacity means I can capture information in whatever way is most convenient at the moment, and the system handles it all.

For daily planning, I use my super high-tech “calendar integration.” I take a screenshot of my calendar and show it to Claude. “What should I focus on today?” It analyzes my existing appointments, identifies open blocks of time, and suggests a realistic plan based on priority and available time. I didn’t teach it to do that, it just does it.

The beauty is how contextual it can be. On normal days, I might ask for a structured day plan optimized for productivity. But when I’m not at my best, I can say, “Today I’m feeling horrible and have no energy. What are just the critical tasks I need to accomplish to not fall behind?” Try getting that kind of adaptive planning from Asana.

The system evolves with me, not because Claude has a long memory (it doesn’t), but because I fine-tune the instructions over time. If there’s a behavior I don’t like, I modify the instructions. If I think of a better way to structure the lists, I update the prompts. Just as I would give corrections to an employee. It’s infinitely malleable with minimal effort.

Why Traditional Tools Failed Me

I’ve finally realized why I abandoned every other system. It wasn’t just the maintenance burden, but how the system itself became a task and a form of procrastination.

Traditional PM tools become a job unto themselves. You spend so much time managing the tool that it defeats the purpose of having one in the first place.

Think about the typical workflow:

1. Create a task
2. Assign it to a project
3. Add tags or categories
4. Set priority levels
5. Add due dates
6. Move it between columns as status changes
7. Update fields when things change
8. Archive or delete when completed

That’s eight steps of overhead for each task! I experienced this nightmare when I tried to implement Tiago Forte’s Second Brain methodology. I became so obsessed with structuring everything perfectly that I accomplished nothing of actual value.

With my Claude system, I just talk. The AI handles the organization, categorization, and maintenance. I never have to “clean up” my system or reorganize my boards. The burden of structure falls on the AI, not on me.

The Five Principles That Make It Work

After building this system and seeing it succeed where others failed, I’ve identified five core principles that explain why it works.

1. Embrace Unstructured Data

My thoughts don’t arrive in neatly categorized packages. They come as messy, interconnected streams. This system accepts my thoughts exactly as they naturally occur and handles the structuring for me.

When I wake up with a random idea at 6am, I don’t need to figure out what project it belongs to or what priority level it deserves. I just record a quick voice note. “Just realized we should add a capability assessment phase to the client onboarding process.” Claude handles the rest.

2. Minimize Interface Complexity

The entire interface is a conversation. There’s no learning curve, no buttons to figure out, no fields to complete. The simpler the interface, the more likely I am to actually use it.

I’ve tried systems with dozens of views, filters, and dashboards. They impressed me initially but became overwhelming quickly. My Claude system has just two outputs. My Active Tasks document and my Completed Tasks document. That’s it.

3. Avoid System Maintenance Burden

The death knell for any productivity system is when maintaining the system becomes a task itself. My Claude system requires little to no maintenance from me.

I never need to clean up tags, archive old tasks, or reorganize my projects. I never have a backlog of “task management” to do before I can start actual work. The system handles its own maintenance. (For the most part)

4. Prioritize Backend Intelligence Over Frontend Features

Traditional tools compete on features. More views, more integrations, more customization options. But features aren’t intelligence.

What matters in my system isn’t a pretty interface but the intelligence behind it. Claude’s ability to understand context, process information within a conversation, and organize it meaningfully. All the complexity lives in the backend, invisible to me, while I interact with a simple chat interface.

5. Natural Language Interface First

We’ve spent our entire lives communicating through conversation. It’s our natural interface with the world. My system leverages this by letting me manage projects through natural dialogue rather than forms and fields.

When I say, “The new client proposal is more urgent than the workshop prep now because they moved up their decision timeline,” Claude understands and reprioritizes accordingly. No dragging cards or updating fields required.

What A Typical Day Looks Like

Here’s how I actually use this system on a daily basis.

1. Morning Brain Dump. I start with a voice note of everything on my mind. Some are clear tasks, others are vague ideas or concerns. Claude organizes them all, adding them to my Active Tasks document in appropriate categories.
2. “High-Tech” Calendar Integration. I take a screenshot of my day’s calendar and share it with Claude. “What’s realistic for today given these meetings?” Claude analyzes my available time blocks and suggests a focused plan.
3. Throughout The Day. As new things come up, I record quick voice notes, take photos of handwritten notes, or screenshot messages containing tasks. “Need to research how Anthropic is using RLHF for alignment.” Claude adds these to the appropriate projects.
4. Task Completion. When I finish something, I simply tell Claude. “Completed the client capability assessment framework.” It moves the item from Active to Completed Tasks.

Weekly Review. Every Friday, Claude helps me review what I’ve accomplished and what’s still outstanding, helping me plan for the following week.

There’s no separate app to open, no context switching required. The system lives where I already work, in my conversations with Claude.

The Psychological Factor

The real reason this system succeeds where others failed goes beyond the technical design. It’s psychological.

Traditional PM tools made me feel judged. Empty task boxes and overdue items created a sense of failure that made me avoid opening the app altogether.

My Claude system feels like a supportive partner rather than a taskmaster. It asks questions instead of demanding updates. The red overdue flags that used to make me abandon ship entirely simply don’t exist here.

What really removes the shame factor is that I don’t need to see all the complexity. It’s like an iPhone. Just a screen and a few buttons, with all the complexity hidden away. I don’t need to see the custom instructions all day. I don’t need to see how things are organized or how the lists are structured. That complexity remains in the backend, and I just interact with a chat window, something I do all day anyway.

The Results

The proof is in the consistency. I’ve maintained this system since January, far longer than any previous attempt at organized task management.

More importantly, I’ve completed more tasks, missed fewer deadlines, and experienced significantly less stress around my workload. The mental energy previously spent maintaining my task system now goes toward actual productive work.

Not For Everyone (And That’s The Point)

This system would probably drive a lot of people crazy. It lacks the visual organization that many find essential. It doesn’t have built-in collaboration features. There are no satisfying checkboxes to tick.

But that’s precisely why it works for me. I’ve stopped trying to force my square-peg brain into the round holes of conventional tools.

The best productivity system isn’t the one with the most features or the prettiest interface. It’s the one you’ll actually use consistently. For me, that meant building something that mapped to my natural way of thinking rather than trying to reshape my thinking to match someone else’s system.

What I’ve essentially built is a personal assistant. If you think about it honestly, if you could afford a PA who remembers everything for you, whom you could call anytime to tell you exactly what you need to do and what’s lagging behind, wouldn’t you pay for that? That’s what most executives have. All those fancy task management systems with graphs and calendars and checkboxes? Those are what you need when you can’t afford a PA.

But with AI, now you can have your own personal assistant. You just need to build it.

This article: AIL 3 — Created using AI with extensive human structure and guidance

About the Author: Santiago helps individuals and organizations harness the power of AI without coding. Through AI opportunity assessments and personalized consulting, he guides clients in finding practical ways to implement AI solutions that transform how they work. Visit AISherpa.me to learn more about working together.