Hack Day - Chat (Khat) GPT

Why did the AI buying assistant start a stand-up comedy career? Because it knows how to find the best "deals" of laughter!

Here is hoping ChatGPT is better at being a buying assistant than it is at writing jokes!

ChatGPT is the hottest topic around the industry and, inspired by a recent tech talk by our CTO, we used our most recent hack day to try and answer some questions about what Kogan.com could use ChatGPT for. Can we train ChatGPT with knowledge of our products so that it can be a buying assistant to our customers? Can we train it with our FAQs and customer support information so that it can help answer customer queries more efficiently? Through a few different experiments, we tried to answer these questions.

So what is ChatGPT?

ChatGPT is an advanced artificial intelligence language model developed by OpenAI that specialises in understanding and generating human-like text. It works by analysing large amounts of text data and predicting the most likely responses or content based on the context provided, leading to highly accurate and coherent interactions.

Where to begin?

The ChatGPT team on our hack day was the largest and involved a good mixture of different types of developers. We split into two main segments - our backend developers focused on training the model with our data and creating an API that could be interacted with. Our mobile developers focused on adding a chat feature to our current app, which would interact with the created API.

Implementation

Training the model

Starting out we investigated which language model would be the best fit for our application. Well-known models like GPT 3.5/4 are best suited for precise answers whereas models like Davinci are better at extended, concise responses. As this was a support-type bot with a general audience, we went with Davinci for a better user experience.

The process of training the bot involved creating a GPT vector index containing all our data using the OpenAI language model we’ve selected. We used the Python lib, LlamaIndex to facilitate this process. We were then able to query our GPT vector index with our questions.

We trained one of the models using product descriptions and product manuals and the second included customer support FAQs and a collection of fake customer support tickets to see if we could train positive interactions into the model.

So, with our models in hand, we jumped in with some questions!!

Immediately we found that general information from our customer interactions with the support team, such as customer name, was being used by the model to assume the user’s name etc. Questions like,

“How can I get a refund on my order”,

Would often come back with responses like,

“Sure , to receive a refund from Kogan…”

The security/privacy implications of a model being trained with conversations were becoming obvious and seemed like a bigger problem for another day, so we had to park the customer support branch of this project for the day.

We focused on the product assistant for the rest of the day, which turned out to be a great decision as we got some great results.

API Setup

In addition to building the model, we also needed to host it somewhere to make it available for the mobile app to consume. Initially, we started building out a lambda function behind an API gateway in AWS to host the application, but in the final hour of our hack day realised that the model wouldn’t fit into the lambda function’s available space, so we quickly pivoted by spinning up an EC2 instance using AWS Lightsail and setting up Flask as our API. It was great to see the team mitigate issues as they popped up like this.

Solution

At the end of the day, the team was able to showcase a working prototype, with our mobile app integrating with the API and returning some interesting responses. The most promising response we got was a question related to safety information contained within one of our product guides.

Q: “Can I use the Estelle Hair dryer in the bathroom?”

A: “No, you should not use the Estelle Hair Dryer in the bathroom. According to the safety warnings, it should not be used in the bath or shower, near a bathtub, washbasin or other products that contain water, and should be unplugged after use and the speed and heat settings should be set as required before use if it is being used in the bathroom.”

Another example of a general question was:

Q: “What options are available in the Kogan washing machine?”

A: ” The Kogan 7kg Top Load Washer offers a range of options, including 7kg clothes capacity, 6 wash programs and memory function, 8 water levels and 9 types of water power, LED display and delay start function, fuzzy logic control and detergent dissolve tech, child safety lock feature, and a drain pump filter to avoid blockages. It is made from high-quality, clear ABS material, making it easy to maintain and durable for everyday use. It also features personal preference settings, allowing you to customise your washing experience.”

At times, the model wouldn’t align the correct information with a given product but this was always good for a laugh.

Q: “Which Kogan toaster would best suit my family of 6”

A: “The Kogan 4 Slice Toaster (Silver) would best suit a family of 6. It features two independent browning controls, seven adjustable browning settings, built-in Defrost and Reheat functions, an ultra-sleek silver design, a weight capacity of 50kgs, and a frame material of wood. It also has a removable crumb tray at the bottom for easy cleaning.”

To improve this we could have provided our data in a more structured format, but for this experiment all our data was unformatted.

One of the main problems we discovered was the time it took to get a response from the API. Responding often took 5-10 seconds to come through and would come through as complete responses rather than each word like the ChatGPT client in your browser does. This was just a limitation of the library implementation and setting, so could be mitigated with a more bespoke solution.

As an added bonus, our mobile developers also took the time to work with SwiftUI, with one even using ChatGPT to assist with this along the way, and we were all impressed at what it was responding with.

Final Thoughts

Overall it was definitely a successful hack day for the ChatGPT team and highlighted the immense potential of the use cases of ChatGPT at Kogan.com. The hack days are a great opportunity to collaborate with other squads within the engineering team which we wouldn’t usually interact with on a daily basis.

MARCH 2022 HACKDAY (Part 1) The Morse Man bot

.. - ... / .... .- -.-. -.- / -.. .- -.-- / .- - / -.- --- --. .- -. -.-.--

The above elegant line of dits and dahs is Morse code. Developed in the 1800s, it has stood the test of time, and with its combination of nostalgia, intrigue and retro vibe, it’s primed for a resurgence in popularity. Instant messengers can feel a bit synthetic and for those who long for a more analog, authentic communication, Morse code may well meet this need. As part of our regular hack days at Kogan.com, and with the intention of improving Slack by adding personalised features, a Morse code bot was just one of a plethora of bots that was created on the day.

So how did we achieve this greatness?

It started with a repurposed Google Cloud project and a quickly assembled Flask app to lay the solid foundations on which to build. Because we care about developer experience, we added UAT builds on push to our dev branch, and because we care about consistency, we set up some pre-commit checks. Once that was done, it was time to start with the meat and potatoes, and get some functionality working. After examining Slack documentation we knew vaguely what we needed. That was to create a custom Slack app bot and use slash commands to access our Flask app. Essentially, one can access a Slack app bot by typing “/bot-name” into the message box and whatever message you write after this is sent as a parameter to the request url specified. The response from this request is then posted in the channel from which it was sent. We could now start work on the individual bots.

The Morse Man bot:

Our progress so far had worked well, but we soon stumbled upon a constraint that was rendering the Morse bot worthless. The problem is as follows, should you wish to utilise the Morse bot to encode a message, the bot replies to the channel with both of the following messages:

Morse Man Slack bot posts:

@userTryingToBeSecret sent this message: /morse Here is my super secret message

Morse Man Slack bot posts: .... . .-. . / .. ... / -- -.-- / ... ..- .--. . .-. / ... . -.-. .-. . - / -- . ... ... .- --. .

No point sending a secret message if you send the real message with it at the same time. Concerned that this could affect uptake and not wanting to jeopardize the project, a better technical solution would be necessary. Further investigation found there is an option to set the bot response to only be visible to the sender, and it was also found the Slack API can be used to send messages ad hoc. Combining these 2 possibilities we are now able to do the following:

• Use a slash command to call the bot
• Send a request to the Flask app (and return no response)
• The Flask app receives the message as a parameter, and using sophisticated algorithms encodes it into Morse. At the same time we get the channel ID from the request API
• Using the separate Slack web API, make a request to the channel and send the message!

Lastly, our Slack bot could only be taken seriously if it has a legitimate avatar, and hence our UI/UX designer got involved to add a polished feel to the project. The end result can be witnessed below.

The Horse Man bot:

Following on from the development and good learnings of the Morse Man bot, but instead of converting into Morse code, it uses an alternative proprietary encryption method to convert into horse dialect. Also complete with its own avatar.

Impersonator bot:

A skunk works project that was developed in the shadows. Someone noticed that you could supply any name and icon with the API request, and this essentially meant you would be able to impersonate any user on Slack(this is true apart from the small word ‘app’ after the username indicating its true source) By using the Slack API that returns user details, the bot can accept a username and a message, then automatically retrieve the users Icon and repost the message as that user. A dangerous discovery that was essentially banned immediately.

Humor bot:

Using the same foundations, but then uses external APIs to fetch jokes from Chuck Norris and Ron Swanson to inject some humor where required.

Dashboard bot:

A truly value adding bot. The idea was for a bot to provide snapshots of our Grafana dashboards when requested on slack. This could provide quick access for everyone and also add value if needed as part of a discussion. It leverages Selenium to login and render the dashboard, as well as take a screenshot, which is then returned with the Slack web API.

Unmorse bot:

Decodes Morse code. This has proven to be particularly useful, as it turns out not many people can read Morse code directly.

We now have a bunch of nifty bots at our disposal and a better understanding of the functionality available for integrating with Slack. At Kogan.com, hack days are considered an important tool in giving us the time to pursue alternative projects of our own choice. The skills and ideas seeded from a hack day can result in something that adds direct benefit or can be an opportunity to learn and upskill ourselves.  It also adds to the culture of our workplace and supports team work by giving us the chance to work with people outside our normal team.