Crispin Courtenay

In January of this year I decided to create a sophisticated WordPress based website based on an Amazon Web Services (AWS) cloud network. There was a twofold reason to do so. First the site was to act as proof of concept for a side project that I am working on, Fermented.Media. Which was created to provide localized reporting on the wine, beer and distilled spirts regions of California, to tourists and locals alike. It’s also a business specifically crafted be a part-time venture, as to not… Show more

In January of this year I decided to create a sophisticated WordPress based website based on an Amazon Web Services (AWS) cloud network. There was a twofold reason to do so. First the site was to act as proof of concept for a side project that I am working on, Fermented.Media. Which was created to provide localized reporting on the wine, beer and distilled spirts regions of California, to tourists and locals alike. It’s also a business specifically crafted be a part-time venture, as to not interfere with any projects that I am working on.

Secondly, building this project on AWS allowed me to take a deep dive into all the relevant technology that Amazon has on offer and learn by doing—and breaking things—in a sandbox environment where the worst possible outcome for failure is spending four minutes to fire up a new server. Thousands of pages of whitepapers later, and more false starts than I can count I was able to create a resilient, highly available WordPress multisite. While it still has some performance tuning to implement, its speed is staggering, with a 22 millisecond server response time (10.6 milliseconds being the speed of light between the testing server and the closest AWS data center). When fully tuned I am expecting sub 2 second page load times on graphically rich pages. Perhaps most impressively, the entire AWS network costs around $100 per month at a development level, but has the capacity for almost unlimited scalability with the push of a few buttons.
Show less

This is a semi-intelligent bot designed to support winery sales and service. WINEBOT interfaces with 3rd party POS and e-commerce and event sales systems. It is designed as a PaaS and will be on market shortly.

A highly-experimental technique for validating machine-learning and artificial intelligence results on mission-critical tasks, where the cost of being wrong can cost a company’s future, or somebodies life.
Separate AI instances in parallel and use an overarching AI to manage them, utilizing Monte Carlo Game Tree Simulations. The more the AI “plays” the greater it’s chance of “winning”. If there is correlation or near neighbor results on two or more AI’s then it is time to take notice of… Show more

A highly-experimental technique for validating machine-learning and artificial intelligence results on mission-critical tasks, where the cost of being wrong can cost a company’s future, or somebodies life.
Separate AI instances in parallel and use an overarching AI to manage them, utilizing Monte Carlo Game Tree Simulations. The more the AI “plays” the greater it’s chance of “winning”. If there is correlation or near neighbor results on two or more AI’s then it is time to take notice of these results. If the majority of AI’s agree on something, it is treated as fact.
Show less

This Virtual Agent is a working proof of concept for user self-service in a complex automated help desk environment. Ubuntu Bot is a trained AI drawing from nearly ten million Question and Answer sets, and all official documentation. The bot looks for the evaluates the best possible answers for any question about Ubuntu, and weighs the results with such factors as author reputation, upvotes, views, status, the number of similar questions, nearest neighbor and actively predicts what the users… Show more

This Virtual Agent is a working proof of concept for user self-service in a complex automated help desk environment. Ubuntu Bot is a trained AI drawing from nearly ten million Question and Answer sets, and all official documentation. The bot looks for the evaluates the best possible answers for any question about Ubuntu, and weighs the results with such factors as author reputation, upvotes, views, status, the number of similar questions, nearest neighbor and actively predicts what the users next steps may be, and prompts the best answers.
Show less

CONCEPT
A brand (Project Tutti Frutti) premium fruit wine made with fresh fruits, exotic herbs and botanicals combined with energizing root ingredients such as ginkgo biloba.

MARKET
Primary: 21-30-year-olds (19 in Canada, 18 in EU) who do not want a normal alcoholic beverage, but something fruity, refreshing and unusual. The primary market is night clubs and bars, however retail distribution could follow on the heels 12-18 months later.

Secondary: The energy drink and… Show more

CONCEPT
A brand (Project Tutti Frutti) premium fruit wine made with fresh fruits, exotic herbs and botanicals combined with energizing root ingredients such as ginkgo biloba.

MARKET
Primary: 21-30-year-olds (19 in Canada, 18 in EU) who do not want a normal alcoholic beverage, but something fruity, refreshing and unusual. The primary market is night clubs and bars, however retail distribution could follow on the heels 12-18 months later.

Secondary: The energy drink and vodka crowd.
Tertiary: People who dislike standard alcoholic drinks.
Show less