Craig Stuntz

Columbus, OH

https://www.craigstuntz.com

My interests are math, compilers, provers, fuzzers, security, formal methods, quantum computing, and homomorphic encryption

Employment

Improving / Director of Consulting Jan 2014 - Current
At Improving, I both consult for our clients and I work to raise the technical excellence of the Ohio enterprise. I am on the leadership team and also do sales engineering
for DOmedia
  • Write Java code to integrate DO’s proprietary buying platform with occasionally-documented APIs from their integration partners.
Java, JavaScript, MariaDB, AWS
for WillowTree (formerly Dynamit)
  • For an auto manufacturer, developed web services in Python/Chalice/AWS Lambda, and identified areas in the code which would be barriers to the business changes they saw coming in the near future
  • For a textbook publisher, built an interactive training system using TypeScript and OpenAI/ChatGPT. Worked to ensure the safety of using a chatbot in this context
  • For a national restaurant chain, rewrote their public web site with TypeScript, AWS Lambda functions, terraform, and C# (Optimizely CMS) code. Established a “security-first” culture within the team
  • For a national convenience store chain, implemented a custom Kafka Connect Sink to replicate data from Kafka to segment.io. Hosted this on AWS EKS with Helm. I actually implemented this twice, which is a fun story too long to include here; ask me!
  • For an international beverage producer, wrote Azure Functions in TypeScript to import data from Salesforce to SQL Server Azure
  • For a major utility, did a complete rewrite of their web site/middle tier after two other organizations had tried and failed, using Spring Boot and Ansible
  • For a national hotel brand, did C# code optimization to fix blocking production issues
  • For a national hotel brand, wrote a Salesforce to Tableau data synchronizer
  • Wrote an Android application and Java server for Village, an employee and customer engagement application
Ansible, C#, Java, JavaScript, Kafka, Kubernetes, MySQL, Python, Scala, SQL Server, terraform, TypeScript, vert.x
for Grange Insurance
  • Wrote F# code which used machine learning and strong types to generate code and unit tests from GBs of XML, producing 80,000 lines of C# code without a single reported production bug
  • Wrote a dashboard for actuaries which combined data from about a dozen different systems, including mainframes and web applications
C#, F#, SQL Server
Spruce Media / Software Engineer Feb 2012 - Dec 2013
  • Analyzed, designed, implemented, tested, and deployed features for Spruce Media’s web-based Facebook ad buying platform
  • Supported new team members in other states
C#, JavaScript, Ruby, SQL Server
Vertex Systems / Application Architect Jan 1999 - Feb 2012
  • Transitioned the company and development team from Win32 legacy applications to modern web products using contemporary technology
  • Wrote a DIBOL-to-DIBOL compiler in Delphi which generated code for Y2K dates, meeting an “impossible” deadline and saving the company
C#, Delphi, InterBase, JavaScript, SQL Server
Wayne State University Department of Physics and Astronomy / Programmer 1994-1995
  • Wrote computer software for statistical analysis and simulation of results from high-energy nuclear physics experiments at BNL’s AGS accelerator
  • Translated scientific papers on nuclear physics into executable code
Fortran!

Speaking

Please see my list of current presentations (https://www.craigstuntz.com/presentations.html) for full details, slide decks, and video, but here is a selected list of talks I do:

Selected Projects

Checking Dependent Types with Normalization by Evaluation (Swift Version)

https://github.com/CraigStuntz/bidi/

A translation of David Christiansen’s bidirectional typechecking Haskell example into Swift.

aural

https://github.com/CraigStuntz/aural/

A command line utility, written in Swift, to manage Audio Units on macOS

Fizil

https://github.com/CraigStuntz/Fizil

A fuzzer for .NET applications written in F#. Does in-memory instrumentation of unmodified binaries.

TinyLanguage

https://github.com/CraigStuntz/TinyLanguage

An educational compiler, demonstrating lexing, parsing, AST generation, type checking, optimization, and code generation, written in F#.

Fantam

https://github.com/CraigStuntz/Fantam

F#, purely functional port of Bob Nystrom’s Bantam Pratt Parser example.