Books and Resources on AI Engineering

Apart from staying up to date through websites there are a number of good books on AI Engineering. Here is a recommended reading list.

AI Engineering, Chip Huyen (2025, O'Reilly) - really good book on building systems on top of LLMs. Chip's Github is here.

Hands-On Large Language Models, by Jay Alammar and Maarten Grootendorst (O'Reilly) - uses Python to convey an understanding of how LLMs operate under the hood, covers similar ground to AI Engineering - definitely worth reading. It has quite a few text processing canned examples which are quite interesting.

Mathematics for Machine Learning, by Deisenroth et al. - not as directly connected to AI Engineering but good at explaining some of the underlying maths of ML intuitively (and in somewhat long winded fashion - at least from an engineering perspective).

Books and Resources on Software Design

So you want to become a legendary Windows software designer? Why not try reading some Unix first?

The design of operating systems can yield a rich source of knowledge and insights into software design. Take a look, for example, at:

The Design of the Unix Operating System, by Maurice J. Bach, first published in 1990, is a great book on how the Unix kernel has been designed and written, starting with an overview of the file and process subsystems, files and processes being core concepts in any Unix system. 

The beauty of this book is that it teaches you to think of the subsystems you need to create within the larger architecture of a software application in order to ensure conceptual scalability.

The book is also available on GitHub.

Good Books on Active Directory, Network Operating System and Cloud Computing

Active Directory, 5th Edition, May 17th 2013, O'Reilly, Brian Desmond, Robbie Allen et.al. - recommended reading. A great book on all things Active Directory.

For wider reading on network operating systems, and how they might change with the cloud, Barrie Sosinsky's books are good ones to dive into:

Networking Bible, Barrie Sosinsky, September 2009, John Wiley and Sons (almost a thousand pages of networking goodness)
Cloud Computing Bible, Barrie Sosinsky, January 2011, John Wiley and Sons

Recognised Security Tomes and Resultant Concepts

Recognised security tomes include:

Bruce Shneier's Applied Cryptography

and "The Rainbow Series" which include requirements for systems that process sensitive information:

The Orange Book - nickname for the TCSEC (Trusted Computer System Evaluation Criteria), mandates a Security Policy to be defined for the computer system. The security policy must be "explicit, well-defined and enforced".

An interesting concept that the Orange Book refers to is the distinction between Mandatory and Discretionary Access Control. Mandatory Access Control impose limitations on "subjects and objects" whereas Discretionary Access Control is more focused on individual user access needs (and access control based on the "groups" to which they belong). The Orange Book doesn't discuss specific implementation abstractions of these concepts though.

The Rainbow Books came out of a recommendation by a Task Force set up in 1967 to analyse computer security safeguards to protect sensitive information. These concerns relate not just to building secure systems, but evaluating and auditing them as well. The Orange Book addresses these dual concerns.

There is also a set of Compact Disc format specifications also known as "The Rainbow Books" (including the 1988 CD-ROM format specification, 1993 VCD Specification).

SQL Server Security - Unleashed

What you Will Learn

What server principals are
What types of server principals there are
What schema to query to access the "principals" list!

The Low Down on SQL Server Security

An average Windows Joe needs to know something about databases, and SQL Server databases in particular. But you can't claim to be a database know-it-all without knowing "un peu" of SQL database security, in particular the concepts of Principals and Securables.

A great book to accelerate your learning about SQL Server Security is "SQL Server (2008 R2) Unleashed" from SAMS Publishing (famous for their "Unleashed" series and "How To" books). The concepts are well-presented and can save you time before diving into specific details strewn across the MSDN jungle.

Speaking of concepts, let's introduce the most basic ones; requestors of resources, resources and permissions. Turns out, these concepts have a rather different terminology in the SQL Server worldspace.

1. A Requestor of a SQL Server Resource is henceforth known as a Principal.  These principals may be Windows users, SQL Server users and so forth.

2. A SQL Server resource is henceforth known as a Securable.

3. Permissions link Principals with Securables.

But you can't know-it-all about SQL Server databases without having some tools to practice with. For this, the SQL Server 2008 R2 Management Studio is an excellent choice.

Some simple queries to get you "synced" with the SQL Serve Security spirit:

1. select * from sys.server_principals

This so-called "catalog view" in SQL Server (nothing more than a "window on metadata") identifies all "server-level" principals and what type they are; an example would be the "sa" account which has type "SQL_LOGIN" (and incidentally cannot be removed). Other types include WINDOWS_LOGIN (self explanatory!), WINDOWS_GROUP and SERVER_ROLE. All are types of PRINCIPALS in a SQL Server database.

However, there is another, less documented type of "server-level" principal, the CERTIFICATE_MAPPED_LOGIN. Some systems use digital certificates as an additional means of authenticating users.

Good Books to learn Microsoft Web Technology

"ASP in a Nutshell" (published in 2000 by O'Reilly) is a good introduction to the "pre-.NET" web programming era. Certainly it is worth reading just to know the concepts that led to more fancy ASP.NET and all the other technologies that followed.

"Professional IIS 7.0" (published in 2008 by Wrox) is an awesome introduction to the new features in IIS 7.0 and what sets this ground-up rewrite apart from previous versions of Microsoft's flagship web server product.

What is Software Science? Let's ask Sedgewick.

What's the latest book on software science you have recently read?

Wait - first off, what IS software science?

If you had to teach a course called software science to cover the more scientific aspects of modern computing, what would it include? Probably the following:

• basic algorithmic knowledge, like how to analyse an algorithm (time complexity and space complexity - yes this is still important, e.g. when programming for small devices) and the classic algorithms like quicksort and bubblesort, what their characteristics are
• basic knowledge on audio compression technology like MP3 and Ogg Vorbis and the underlying concepts
• working knowledge of maths underneath RSA and other cryptographic protocols and algorithms (prime numbers, number theory)
• basic computational physics for use in programming physics engines and ray tracers used in game development

What are the "great tomes" of software science?

Consider the first category of "basic algorithmic knowledge". Robert Sedgewick's 1983 tome simply-named "Algorithms" is always a firm favourite with students, and even covers a bit of linear programming as well (albeit under "Advanced Topics). There is also a fantastic book called "Computer Algorithms C++" by Ellis Horowitz (University of Southern California) and Sartaj Sahni (University of Florida) and Rajasekaran Florida) published in 1996, which delves quite nicely into parallel algorithms also, which is getting more fashionable with the whole Google map-reduce malarkey.

Caveat: Sedgewick says you probably need about a year's worth of programming experience to get the most out of his book.

In terms of audio and multimedia technology (one of the areas of computer science strongly influenced by maths and more traditional engineering disciplines, like electronics engineering) most of the information can be gleaned from the web but there are of course books available too. One such is Digital Audio Signal Processing, Udo Zolzer, Wiley, 2008 (the book has been described as "useful" but not very accessible).

For RSA, Shneier's Applied Cryptography is a must (highly praised by Dr. Dobbs Journal when it first came out). For computational physics, try: David Eberly's Game Physics, 2003. To get the most out of this book, you really need to get psyched about mastering Newtonian and Lagrangian dynamics.

What is Secure Coding?

UNMANAGED CODE IS THE ROOT OF ALL EVIL (LET THE CANARIES PROTECT YOU)

The issue of "secure coding" comes into play when you are writing "Internet" software.

An old classic example of "insecure code" would be buffer overruns (something that can occur in poor C++ coding). It's rather harder to do in Java or C# due to run-time checking of array boundaries. Michael Howard and David LeBlanc call buffer overruns "Public Enemy Number One" in their book, "Writing Secure Code" (published by Microsoft Press, December 2004), also known as "The Grey Book".

Visual C++ .NET comes with a /GS option that employs canaries to neutralise buffer overflow attacks (its etymology comes from the "canary in the coal mine" analogy). The canary is a known word inserted into the stack to monitor buffer overflows, such as if data in the stack gets corrupted, the first data to get corrupted is the canary.

One of the "good practices" identified in The Grey Book is identifying parts of your infrastructure in C or C++ which could be re-implemented in a managed language, to reduce the risk of classes of attacks, such as buffer overflow.

GREAT THING ABOUT C#

One of the great things about C# is you don't have to worry about the so-called "secure coding" issues that you have to worry about (or at least be mindful of) in C++.

PRINCIPLE OF LEAST PRIVILEGE

This is a useful security principle, whereby you give a user the rights to do what they need to do and nothing more.