My Personal Knowledge Management (PKM) Workflow

(Photo by Ash Edmonds on Unsplash)

Ever since I was a child, I have had a special liking for notebooks. I like to record some numbers, temperature, time, how many pigeons in the building next door outside the window, how long it takes to fly back to the pigeon loft, etc.

As I get older, I have always regarded notebooks as a tool for recording, memo, and quick review. I always feel that it does not match my workflow, whether at work or at home. Until the last few years, I have been more coherent, and recorded my current personal knowledge system (PKM) workflow. This workflow is a process that I understand and generalize to my personal situation from various sharings from people who are smarter than me and more focused on PKM, Smart Notes, etc. My process may not be suitable for everyone, but I hope that by sharing ideas, I can bring out some dialogues and discussions, so that we, and even our next generation, can reduce the time of groping and improve learning or work efficiency. And then make some contribution to society or human beings.

I disassembled the “personal knowledge system” and divided it into three parts: “personal”, “knowledge” and “system”:

  • Personal: Use it for yourself, as long as you are comfortable, don’t stress.
  • Knowledge: Knowledge is knowledge with confidence about the act of “recognition” and “recognition” of a subject, and that knowledge has the potential to be used for a specific purpose use. 1
  • system: there are processes, tools, and possibly automation.

People who are forgetful are always forgetful (oh don’t get me wrong, it’s me), by incorporating context, and designing various triggers, to allow myself to start work tasks. Combined with loop inspection, an inner cycle is formed. With goal orientation, form external output.

I expect (hope) to go back and update this article incrementally, some tools may be tweaked over time, and this article will consist of 7 process steps:

  • 1️⃣ Input sources
  • 2️⃣ Tools for processing input sources
  • 3️⃣ Inbox
  • 4️⃣ Scoring mechanism
  • 5️⃣ Distilling
  • 6️⃣ Executing
  • 7️⃣ Expressing

I cut it relatively finely, you can use it as a reference, and then call out the process steps that are suitable for your own use.

Two paragraphs are attached at the end of the article:

I hope it will be helpful for you to build your own PKM.

“Use it for yourself, just feel comfortable, don’t stress.”

1️⃣ Input sources

In the process of cross-field self-study, you will encounter various sources of information in various formats. Depending on the individual’s situation, some forms of information sources are easier to absorb and others are not. Some can be designed to automate a series of workflows, some are more structured, and some are more fragmented. Depending on our own needs, we make adjustments on a regular or irregular basis.

Below are some input source types that I commonly use. You can also observe various aspects of your life and sort out your commonly used ones.

graph LR; B(Books) --> IS(1. Input sources); A(Articles) --> IS; V(Videos) --> IS; P(Podcasts) --> IS; C(Courses) --> IS; CT(Conversations and thoughts) --> IS; class IS mermaidTier1; class B,A,V,P,C,CT mermaidTier2;


Books are mostly more structured sources of information{, and relatively solid content, verified content, organized content.

Of course, there will also be some books, which are actually a collection of multiple articles, which will be handed over to the “scoring mechanism” in the following chapters for screening, or select some paragraphs to read.

Under what circumstances would I choose to read books:

  • Most books are fairly cheap. (Don’t get me wrong, the absolute value of the selling price may be unaffordable compared to the current state of each of us, but compared to the effort, time, and resources invested by the author, it can be sold at such a price. relatively complete knowledge content, most of which are quite cheap.)
  • In the selection of books, I will mostly choose classic books that have been published for a period of time, so that there will be more people who have already screened and even analyzed for me. In the historical evolution after the publication of books, which ones can withstand verification, which ones are not, whether new parameters need to be added to make judgments, and so on.
  • Combining the above two points, if you obtain it in the form of e-book, the cost of acquisition is often cheaper, and there is an opportunity to connect the digital workflow for note collection.

PKM Automation:

  • Currently Safari Books (O’Reilly Media), Kindle can be imported to Readwise for compilation .
  • Kobo: I’m still in the process of adjusting.


Articles refer to content that can be read in a short period of time, such as 15 minutes or less. Provide relatively fragmented, real-time information or knowledge, which usually requires subsequent processing, organization, and linking to convert it into easy-to-use, ready-to-use knowledge.

At present, I will stream the articles collected from various places into Instapaper before reading (of course ,there will be situations where some content provided from local business does not comply with the HTML semantic structure and the content or structure is missing, but usually articles on such platforms can usually be skipped and give them lower score later on, or continue to encourage yourself to read similar articles in other languages).

Under what circumstances would I choose to read articles:

  • Use fragment time (waiting, waiting for people, waiting for transportation, waiting for food, etc.)
  • The cost of obtaining is low (but mining is expensive, so it needs to cooperate with the score mechanism as a source screening to improve conversion efficiency)

There are many ways to obtain articles. I use Instapaper that can easily collect the URL of an article and read it later on various platforms (computer browsers, various mobile apps), and further adopted Instapaper Premium, mainly using its “Unlimited notes” function, and then imported from Instapaper to Readwise for the aggregation.

If I can’t put the URL into Instapaper in some cases on my phone, I will instead put it into the ToRead Inbox of a todo tool temporarily.


Watching videos to learn new knowledge always makes people feel that it is easy and easy to get started. However, from the perspective of time conversion efficiency, it may not be the best solution. (Some of my friends can only learn by watching videos, and with some tips, they also achieve quite rich results, so please don’t care about other people’s best solutions, as long as you have the intention, try to find your own best solution.)

Under what circumstances would I choose to watch videos:

  • Situations that require picture explanations, video explanations , such as learning recipes (seeing a dozen different chefs at a time interpreting a dish or ingredient in different ways); new software screen operation (for example, I have been learning to set thinkorswim’s software operation screen), how to plant plants, etc.
  • Suitable for learners of image memory, and at the same time deepen the memory impression through three channels of image, sound, and even subtitle text
  • Time to relax
  • You can adjust the playback speed of the video accordingly.


Podcasts are input into the brain in the form of audio, and most of the time, you can free your hands to perform different tasks at the same time. The usual tasks are to receive information without using audio. My current listening rate is low, mostly because of the low density of information (listen to a clubhouse with a relatively short time to introduce yourself?!), but I still leave it to the “score mechanism” to help filter out sources.

Under what circumstances would I choose to listen to podcasts:

  • I can free my hands for other tasks, but I can’t listen to podcasts during cooking, I need to listen to the sound in the kitchen
  • More relaxing moment
  • Collect some stories

If you want to take notes while listening to podcasts with rich content, I currently use the Airr app on my mobile phone to listen to the podcast content, and hear the clip you want to take notes on. You can press the headphone function key of “Previous Song”, Airr will capture the last 45 seconds of audio clips, called “AirrQuotes”, by default. After listening to the entire podcast, then organize these snippets and enter text notes. Then you can concatenate and import to Readwise for integration.


In fact, I want to arrange the importance of the courses right after the books, but after thinking about it, I think I should first introduce the articles, videos, and podcasts (audio) in more detail, and then talk about the courses.

Courses is often a process and environment in which multiple senses are integrated and presented to learners.

In high school, I came into contact with the curriculum structure design of Project Adventure (PA) Exploring Education in the open school spirit, which opened my door to curriculum design, speeches, etc. The photography club, the extension of elective courses to film and stage, and the opportunity to organize a lot of materials and practice speeches during AIESEC have always made me feel inexplicable about teaching and courses.

The source of the course is similar to that of books. The lecturer (author) has accumulated a lot of time for research on a certain topic or knowledge area, and then crawls out a way to present it to the learner (reader). Using this ruler as a measure, it can be slightly classified, which are worth investigating, which are assembled, and which are assembled. The other rulers may be left for the “score mechanism” to discuss.

Under what circumstances would I choose to attend courses:

  • Like books, structured and systematic knowledge structure can greatly save your time in exploring. (The costs are the same as the books, which are pre-screened.)
  • Compared with books, the courses have more interactions with lecturers, and interactions with classmates. Short-term interaction is to solve current problems or doubts through question and answer, and in medium and long-term interaction, you may make good friends who communicate with each other for a lifetime.

Courses are handled similarly to book notes, basically divided into physical and electronic forms.

Conversations and thoughts

I categorize most sources that are entered within 5 seconds into this category, including formal meetings, informal meetings, social media (Facebook posts/comments, Twitter tweets, Instgram posts, etc), etc., most of them are snippets of information, some of which require post-processing and reorganization, and this ratio usually varies from person to person.

Because conversation is mentioned here, there is actually a part of output, and output is related to linking and creation, and the output part is reserved for later chapters.

How to deal with such input sources:

  • Most (one sentence, or shorter sentences) I will temporarily store in my daily notes. (belongs to fleeting notes)
  • If other people’s insights have passed the “score mechanism”, they will be temporarily stored in insights notes. (belongs to fleeting notes)

How can you get started

(To be added)


Q: How did you start designing the PKM workflow?

A: As the organization grew and the number of simultaneous projects at hand increased, as a technical manager, I was constantly learning new things at the same time, thus reaching new heights of confusion. Therefore, I began to observe how others deal with a large amount of data across contexts and fields, and then settled down to sort out and define my own workflow. At the beginning, it was quite tupao (self-made, simple but works). Later, when I found that there was a keyword PKM, I searched all the way. After dismantling various fragments and steps, my own framework gradually emerged.

The simplified version looks like this:

graph LR; IN(Input) --> P(Process); P --> OUT(Output); class IN,OUT mermaidTier1;

After adding context and Inbox, it looks like this:

graph LR; C1(Context) --> T1(Input Trigger); T1 --> PP1(Pre Process); PP1 --> IN(Inbox); PP1 --> T1; IN --> PostP1(Post Process); PostP1 --> Out1(Output); PostP1 --> IN; class T1,Out1 mermaidTier1; class IN mermaidTier2;

When I first started to match a variety of different contexts, the information explosion state began to appear one after another, making people feel that they were out of control, but continuing to try to sort out a set of PKM workflow for their own use can keep the complexity of the whole thing under control, and still operate.

graph LR; C1(Context #1) --> T1(Input Trigger #1); C2(Context #2) --> T2(Input Trigger #2); C3(Context #3) --> T3(Input Trigger #3); T1 --> PP1(Pre Process #1); T2 --> PP2(Pre Process #2); T3 --> PP3(Pre Process #3); PP1 --> IN(Inbox); PP1 --> T1; PP2 --> IN(Inbox); PP2 --> T2; PP3 --> IN(Inbox); PP3 --> T3; IN --> PostP1(Post Process #1); IN --> PostP2(Post Process #2); IN --> PostP3(Post Process #3); PostP1 --> Out1(Output #1); PostP1 --> IN; PostP2 --> Out2(Output #2); PostP2 --> IN; PostP3 --> Out3(Output #3); PostP3 --> IN; class T1,T2,T3,Out1,Out2,Out3 mermaidTier1; class IN mermaidTier2;
Loading comments…