Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Doing more with Static Sites + Transcription and Sentiment Analysis Pipeline for Call Centre Recordings

176 views

Published on

2 SPEAKERS IN A SINGLE SERVERLESS MEETUP: 1st) Bhavana Srinivas (https://twitter.com/bhavana1110), Solutions Engineer at Netlify.com; (2nd) Mark McQuade (https://www.linkedin.com/in/mark-mcquade/), Cloud Solution Architect at Onica.com

1. Bhavana showed us how to bring alive static sites with Functions. She dived deeper into building sites the JAMstack.org way and demoed how to hook up these sites with Serverless Hello Worlds to more useful 3rd party services. (recording at https://www.youtube.com/watch?v=l0oWGGctKZc)

2. Mark gave an AWS "This is My Architecture"-style talk, about stitching Serverless components to implement Audio Transcription and Sentiment Analysis Pipeline for a Call Centre client who was exploring the benefits of AI & Machine Learning (ML) in the Cloud. (recording at https://www.youtube.com/watch?v=l0oWGGctKZc&t=2310s)

P.S. Special thanks to Myplanet (https://www.myplanet.com/) for providing the space, and Thundra (https://www.thundra.io/) for providing pizza and refreshments!

P.P.S. If you'd like to speak at any of the upcoming Serverless Toronto User Group events, our Slack community (via http://slack.ServerlessToronto.org) and add your topic to the #want-to-present channel.

Published in: Software
  • Be the first to comment

  • Be the first to like this

Doing more with Static Sites + Transcription and Sentiment Analysis Pipeline for Call Centre Recordings

  1. 1. Tuesday, Mar 26, 2019 1. Intro & Activity Update 2. Community Open Mic 3. Bhavana Srinivas, Netlify: "Doing more with Static Sites" 4. Mark McQuade, Onica: "Transcription and Sentiment Analysis Pipeline for Call Centre Recordings“ 5. Networking 1 ServerlessToronto.org Meetup Agenda
  2. 2. Manning Publications 2019 giveaways: 1. www.manning.com/books/serverless-applications-with-nodejs 2. www.manning.com/livevideo/production-ready-serverless 3. www.manning.com/livevideo/production-ready-serverless 4. www.manning.com/livevideo/serverless-applications-with-AWS 5. www.manning.com/livevideo/serverless-applications-with-AWS 6. www.manning.com/books/serverless-architectures-on-aws 7. www.manning.com/books/http2-in-action 8. www.manning.com/books/event-streams-in-action 9. www.manning.com/books/the-design-of-everyday-apis 10. www.manning.com/livevideo/graphql-in-motion 11. www.manning.com/books/voice-applications-for-alexa-and-google-assistant 12. www.manning.com/livevideo/machine-learning-for-mere-mortals 13. www.manning.com/books/classic-computer-science-problems-in-python 2
  3. 3. ● A single pane of glass for testing & debugging & monitoring your serverless stack. ● Discover issues in serverless architectures visually. ● Test and make sure to deploy to production on Friday 6PM. ● Analyze performance issues and pinpoint bottlenecks. ● Discover root cause of errors and latencies ● Gather performance metrics for your serverless transactions with distributed tracing ● Add as a Layer and instrument with no code change. ● Support for Node.js, Python, Java, Golang, .NET ● Your serverless monitoring data on your favorite platform: Integrate with Splunk, Honeycomb, and ELK Detect, Debug, Monitor serverless architectures with Thundra
  4. 4. Screenshots
  5. 5. ● How Thundra discover the root cause ○ https://youtu.be/jEctnLfbTTs ● Play with the demo: ○ https://demo.thundra.io/functions/default ● Thundra has a free tier: ○ https://console.thundra.io/signup ● Read our blog: ○ https://medium.com/thundra Learn more www.thundra.io @thundraio @thundra info@thundra.com
  6. 6. Community Open Mic 1 10 seconds of freedom to pitch yourself, or your company
  7. 7. Doing more with Static Sites
  8. 8. Courtesy: https://scotch.io/bar-talk/5-reasons-static- sites-rock
  9. 9. JAMstack
  10. 10. Courtesy: https://builtvisible.com/go-static-try- jamstack/ Pre-build your site
  11. 11. But why?
  12. 12. Performance Security Scale
  13. 13. But…. is static enough?
  14. 14. Functions
  15. 15. Functions Workflow Any Git Vendor
  16. 16. Demo
  17. 17. Ease of use
  18. 18. Tools and starting points.. ● Netlify Lambda CLI ● Function Logs ● https://functions.netlify.com/ ● React app + Lambda ● Go Lambda functions ● Event driven functions ● Security for functions ● Customize..
  19. 19. ONICA.COM PREMIER CONSULTNG PARTNER Transcription and Sentiment Analysis Pipeline for Call Centre Recordings
  20. 20. ©2018 ONICAONICA.COM About Mark McQuade • Solutions Architect at Onica • Telephony background for 15 years
  21. 21. ©2018 ONICAONICA.COM About Onica
  22. 22. ©2018 ONICAONICA.COM Contact Center Pain Points • Cost of manual transcription • Its not real-time • Manual process limits the amount of calls that can we QA’d • Simple transcription isn’t enough • Need to recognize black-listed words • Split agent and caller audio • Pulling key words from calls
  23. 23. ©2018 ONICAONICA.COM Customer Use Case • Wanted contact centre audio files transcribed - speech to text • Currently sending to Philippines for manual transcription only - no sentiment analysis being performed currently • Didn’t want to manage infrastructure • Cost was a major factor • Wanted end-to-end solution • Wanted ability to extract key words
  24. 24. ©2018 ONICAONICA.COM The Solution • Amazon Transcribe • Amazon S3 • Amazon Comprehend • AWS Lambda • AWS Step Functions • Amazon Connect
  25. 25. ©2018 ONICAONICA.COM Analysis • Amazon Athena • AWS Glue • Amazon Quicksight
  26. 26. ©2018 ONICAONICA.COM Application Architecture
  27. 27. ©2018 ONICAONICA.COM Application Architecture - Cont.
  28. 28. ©2018 ONICAONICA.COM Alternate Analysis • Amazon Elasticsearch • Kibana for Visualization
  29. 29. ©2018 ONICAONICA.COM Amazon Transcribe
  30. 30. ©2018 ONICAONICA.COM Amazon Transcribe
  31. 31. ©2018 ONICAONICA.COM Amazon Transcribe
  32. 32. ©2018 ONICAONICA.COM Amazon Comprehend
  33. 33. ©2018 ONICAONICA.COM Amazon Comprehend
  34. 34. ©2018 ONICAONICA.COM Amazon Comprehend
  35. 35. ©2018 ONICAONICA.COM Amazon Quicksight
  36. 36. ©2018 ONICAONICA.COM Kibana
  37. 37. ©2018 ONICAONICA.COM Kibana
  38. 38. ©2018 ONICAONICA.COM Worth Mentioning • Amazon Lex • Front end contact center calls • Removes need for manual upfront diagnosis of calls • Add’s a very powerful tool to your contact center work flows.
  39. 39. ©2018 ONICAONICA.COM Get Help - Don’t go it alone https://aws.amazon.com/canada/AWS-Canada-Partners/ mmcquade@onica.com www.linkedin.com/in/mark-mcquade
  40. 40. Join www.ServerlessToronto.org Home of “Less IT Mess”

×