My Gemini Live API based assistant for an experimental CX journey can now read barcodes when conversing with the customer.
1. Customer establishes video WebSocket stream
2. Server stores the last X seconds video frames
3. Model calls tool which sends frames to barcode detector
My ongoing project to detect errors in Genesys flows and chatbots...
It
* Tracks changes to flows/chatbots
* Alerts to sudden increases in errors
* Provides affected Conversation IDs
There's much more I'd love to do, including tracking using Git
#Genesys#Chatbots
I'll be talking in a session at @Genesys's #xperience23 today about optimising bots, here are the links in my talk
Articles from my talk: https://t.co/dXo1yAsWJR
Chatbot tooling: https://t.co/yW0ZGQZbMh
Demo of generative AI testing a Genesys chatbot: https://t.co/HyirzAPBfl
Sketchnote of what I learnt at the first day of @Genesys#Xperience23 (I ran out of time to include it all). The exhibition opened today, and talks start tomorrow!
Customer service bots are essential as consumer use of chatbots continue to rise, but satisfaction is declining. @OVOEnergy software engineer Lucas Woodward joins Genesys experts to dive into how to use automation to build better #bots, faster.
On my way to speak at #xperience23. I've set myself the goal of implementing
ChatGPT into my @Genesys Chatbot testing tool (https://t.co/YDuaWWJ9Hp) during the journey so
anyone with a @OpenAI ChatGPT account can start testing their bots with Generative AI!
I’m on @Genesys's Tech Talks in 20 podcast! I’m talking about automating the development/testing pipelines for chat-bots at @OVOTechTeam, which allows us to reduce the time it takes to deliver functionality to customers and adapt to their feedback
https://t.co/mYQPtRQeiv
A few days ago I posted another article on the @OVOTechTeam blog, this time about a service I created that uses @Genesys Cloud's Participant Data to convey metrics, which my service publishes to DataDog. We can then utilise the power of DataDog with them
https://t.co/Z9AM41swJa
A sketch for a new blog article I'm writing about realtime streaming 'metric' @Genesys participant attributes to @datadoghq. It's currently handling 4.5 million events a month for approx $40 (AWS side) and $20 (DD side) in custom metrics
@peter_senna I usually quickly capture the initial idea on a sketchpad. Then refine it to a more presentable sketch on my IPad. Personally, I find the key is not to add too much detail when refining, just focus on the original sketch at the time of the idea. This is at least what I try to do!
@peter_senna@Genesys Thanks for reminding me. It’s now at https://t.co/AapImonySq
I had the great idea of creating a new website for chat bot specific articles, then decided I’d just use my original site.
I pivoted on the #sideproject to instead expose lower-level metrics that teams might find useful - splitting functionality by Lambdas. I learned last month just how expensive @datadoghq's Custom Metrics can be if you aren't careful about the cardinality of your tags 😱
You can now easily insert multiple @excalidraw library items at once. Select several items (or hold Shift to select a range) and drag (or click) to drop on canvas.
via @dluzar
Try yourself with this library from @SketchingDev → https://t.co/k5td3x3ivB
Think I've come up with a low level architecture: A weekly hit of @jeremy_daly’s serverless newsletter has sustained my interest all things serverless, so I’ve come up with the simplest I can think of to allow me to quickly get something up and running (securely) #sideproject