Erik on Software

Personal Reading List: 1Q2021

My personal reading list for January, February, and March 2021.

Color Key: Special Notes, Completed Reading. Updated throughout the Quarter.

---

At Bat (Schnell)

---

The Model Thinker: What You Need to Know to Make Data Work for You, by Scott E. Page, Basic Books, 2018.

Learning Spark (Second Edition), by Jules S. Damji, Brooke Wenig, Tathagata Das, and Denny Lee, O'Reilly, 2020. Copy provided by Databricks.

Scala for Java Developers: A Practical Primer, by Toby Weston, Apress, 2018. 

---

 At Bat (Langsam)

---

Fluent Python: Clear, Concise, and Effective Programming, by Luciano Ramalho, O'Reilly, 2015.

Practical Statistics for Data Scientists: 50 Essential Concepts, by Peter Bruce and Andrew Bruce, O'Reilly, 2017. Copy provided by MarkLogic.

Scala for the Impatient (Second Edition), by Cay Horstmann, Addison-Wesley Professional, 2016.

---

 On Deck

---

Coding Concepts for Kids: Learn to Code Without a Computer, by Randy Lynn, Rockridge Press, 2020. Copy provided by Amazon.

Python Programming for Beginners: A Kid's Guide to Coding Fundamentals, by Patricia Foster, Rockridge Press, 2020. Copy provided by Amazon.

---

In the Hole

---

Monolith to Microservices: Evolutionary Patterns to Transform Your Monolith, by Sam Newman, O'Reilly, 2020. Copy provided by Nginx.

Presto: The Definitive Guide, by Matt Fuller, Manfred Moser, and Martin Traverso, O'Reilly, 2020. Copy provided by Starburst.

Media Query Source: Part 31

The responses I provided to a media outlet on December 16, 2020:

Media: We are looking for the top reasons for why your MVP may fail or succeed.

 

Gfesser: The top reason for MVP (minimum viable product) success or failure revolves around its very definition. Has consensus been reached amongst stakeholders? Does this consensus include approaching what is to be built like a product? Is the product to be built both viable and minimal? If the answers to these three questions can accurately be communicated as "yes" for a given MVP, it is likely headed in the right direction. If not, the question needs to be raised as to why stakeholders are referring to what is to be built as an "MVP". 

An itemized product backlog first needs to be built. A product backlog essentially defines the product, as it is a list of what is needed to improve the product over time. Products have lifespans, and a product backlog lists the priority order in which each improvement is to be built. When an MVP is defined, it is defined with a subset of these improvements, focusing on what will be minimally needed to get it in front of end customers to start eliciting feedback.

 

However, stakeholders need to understand that while getting feedback is a goal en route to building, deploying, and releasing an MVP, end customers typically do not have visibility to the supporting infrastructure and software beyond the user interface. As such, stakeholders need to understand considerations for longer term viability, making tradeoffs with the builders along the way.

 

Yes, what is built for an MVP can technically be replaced following release, but doing so needs to be carefully considered, especially in cases where the supporting infrastructure, software, and data behind the user interface can be seen as the primary product, as the initial user interface only implements a small number of use cases and the flexibility of the platform is paramount. Certainly, such is the case when application or data platforms are built, as these are intended to subsequently support additional use cases.

 

Lastly, it is beneficial for the team building an MVP to be dedicated to the product and free of dependencies on other groups or individuals, to be cross-functional, and to work in an agile manner. Teams exhibiting these qualities will be self-sufficient and flexible as the buildout takes place. In addition to stakeholder lack of consensus and not approaching the MVP as a product, these are among the common reasons I have seen either failure or extreme challenge.

 

---

See all of my responses to media queries here.

Error in Official AWS Certified ML Specialty Exam Readiness Training

In working through the official AWS Certified Machine Learning Specialty Exam Readiness training, I spotted an error.

Can you see it?  :-)

The formula for accuracy is incorrect. As can be seen at this link, it should actually be presented as follows:

Accuracy = ( TP + TN ) / ( TP + TN + FP + FN )

A few months ago, I presented to work colleagues my path to pass the AWS Certified Data Analytics Specialty exam, and noted coming across quite a few issues in the training I worked through from various vendors, but this error from AWS caught me by surprise.