Site Tools


Thoughts and Tips


Tribal Knowledge

As a staunch believer in Free and Open Source software, I apply the same philosophy to knowledge: by publishing knowledge to as wide an audience as possible, in as accessible and editable a form as possible, the writing gets iterated quickly, improving drastically in quality and relevance.

However, I have encountered many peers who do not share this philosphy; anecdotally, I am told this is out of fear and embarassment of being inaccurate or wrong. A more pessimistic (and presumptious) possibility, is also to use tribal knowledge as contributing to job security; well, your job is secure until the moment you become uncontactable beyond your peers' patience.

Rather than editorializing, I defer back to the facts: The people in my teams who get long-term, reccurent recognition (through the annual review process), are people who document fervently. The people in my teams who get recognized once, are people who are SMEs for a quarter.

Penny Wise, Pound Foolish

One of the weirder things that came out of the tech winter of 2022-2023, is that my managers at Amazon Web Services started to heavily reduce discretionary spending, which went into minutae such as reducing team dinners, or not organizing celebrations of major team successes.

In my role at Amazon Web Services, I was constantly presented with opportunities to save the company anywhere from ≈US$200,000 on an MX960 we didn't need, to millions of US dollars on a rack of equipment, to billions of dollars on datacenter infrastructure, all of which were entirely up to my discretion.

Did we really need to save US$50 on a dinner?


Here are some useful thought experiments and exercises I picked up at work. I have specific mentors that I am very grateful for for these, but I am unsure if they wish to be credited - please write me if you do.


I love retrospectives, because I hate seeing the same mistake made twice. One of the reasons I left my team at Amazon Web Services is an erosion of the rigour we would put into retrospectives, leading, expectedly, to mistakes being repeated between projects; ironically, it was during my time there that I learned most of these retrospective techniques.

Start, Stop, Continue Exercise

This exercise is borrowed from Agile software development, and is pretty straightforward.

The team comes together to ask three questions, and fills up a table with:

Start Stop Continue
What should we start doing? What should we stop doing? What should we keep doing?
work/thoughts_tips.txt · Last modified: 2023/09/17 10:31 by Andrew Yong