Closers final project
1 like107 views
This document describes a capstone project to create a BOTSERVER platform and various conversational bots. The BOTSERVER would allow users to launch and manage multiple bots for tasks like answering FAQs, monitoring conversation tone, and automating meetings. These bots would analyze messaging data through APIs and AI like IBM Watson. The goals are to give people super-human abilities by having bots interrupt with useful insights. Live demos of bots like TONE_BOT and FAQ_BOT are presented built using technologies like Django, Python, and Slack APIs. Use cases include boosting employee and customer service through personalized automated interactions.
Closers final project
- 1. CLOSERS BOTs and Artificial Intelligence to Main Street Byte Academy, NYC March 11, 2016 CLOSERS COHORT - FINAL PROJECT Python Coding Bootcamp Brian McHugh Jared Russell Marc Ira
- 2. BOTS AND AI IN EVERYDAY BUSINESS ➢ DEMONSTRATE TECHNOLOGIES Learned in BOOTCAMP ➢ GO BEYOND Python, Django, JS, CSS/HTML, APIs and CRUD ➢ GET STARTED on CONVERSATIONAL TRANSACTIONS in IM ➢ Learn about ARTIFICIAL INTELLIGENCE applications
- 3. WHY BOTS AND AI?
- 4. GOAL: Enable super-human abilities for everyday people through AI-aided technologies.
- 5. Super-human abilities now REPEAT Freestyle Chess Algorithmic Trading AI-enhanced Decisions
- 6. Harvesting personal/business intelligence is increasingly difficult due to the growth of structured and unstructured data. Chart: Quantity of global digital data 2010 - 2020 The problem Yes, it’s an INSIGHTS ARMS RACE
- 7. Why now? +400% (32 ZB) more data in FOUR YEARS
- 8. Messaging: A Large Sector of the Big Data Flow
- 9. The solution Super-human abilities Enlist artificial intelligence technologies to SIFT through data flow, have BOTS (smart agents) INTERRUPT us with important insights for ACTION. REPEAT.
- 10. Our “BOTS”: ➢ Carry out a complex series of actions automatically ➢ Reside inside digital conversations ➢ Can ‘listen’ in and offer some interaction with the client ➢ Through AI, can learn and adapt
- 11. How it works Step 1 Big Data We cannot absorb 99% of the data we encounter. Machines - smart agents, bots - will analyze our personal and business data feeds. Step 2 Smart Agents (Bots) Need to know a way to interrupt us when they find something of value. Step 3 Interruption When they interrupt, they will learn from our reactions. OVER TIME, they get better at anticipating and assisting us with context. REPEAT
- 12. (Conversation) BOTs are Smart Agents Example: Facebook M Virtual Assistant + UBER In-Message Transaction
- 13. BOTSERVER, Bots & A.I. in Instant Messaging Flows Our Capstone Project:
- 14. BOTSERVER BOT_1: Tone_Bot (AI) BOT_2: FAQ_Bot BOT_3: Super_Kei BOT_4: Billy_Bot BOT_5... BOT_6... Our Capstone Project: BOTSERVER + Slackbots + Watson AI
- 15. The Technologies I. BOTSERVER ➢ Django Framework ➢ Python ➢ Javascript ➢ Node.js ➢ HTML/CSS II. The BOTS (in Slack) ➢ OAuth ➢ API Calls ➢ PostgreSQL ➢ MongoDB ➢ IBM Watson A.I. Platform ➢ Slack API ➢ Howdy.ai
- 16. I. BOTSERVER Step 2 A Website that allows a user to: ➢ Load up Active BOT details from DB ➢ Wrap BOT info in a BOTwrapper object ➢ Choose BOTs (e.g. Tone_Bot or FAQ_Bot) ➢ Startup BOTS connection to Slack ➢ Keep Track of all Running BOTS
- 17. II.1 The BOTS: TONE_BOT (AI) Front End Step 2 A SLACKBOT that: ➢ Listens In on an IM Conversation ➢ Parse Conversation through Watson A.I. ➢ Determine the Conversation Tone (e.g. Anger, Joy, Sadness, Disgust, Fear) ➢ If THRESHOLD is reached, AI will REACT to Dial Down the Tone
- 18. II.1 TONE_BOT (AI): Step 2
- 19. II.2 FAQ_BOT Answers FAQs from Multiple Clients Step 2. Engages potential students . Injects BYTE Bootcamp information to inquirers
- 20. II.3 Super_KEI_BOT CULTURE BOT . Occasionally interacts with students . Injects BYTE Culture nuggets like Friday Checkins, Clean Mugs, TA support, etc...
- 21. II.4 BillyBOT DEV TEAM Standups Meeting BOT . Collects Daily Dev Team Status Info . Saves time by polling each team member simultaneously
- 22. LIVE Demo
- 24. USE CASES: BOTS ➢ FORCE Multiplier for harried employees ➢ TIME Saver for repetitive, ongoing tasks ➢ PERSONALIZED interaction with clients ➢ EASIER ONBOARDING for employees and clients ➢ DYNAMIC Service Delivery
- 25. USE CASES: BOTSERVER ➢ ALLOW Wide Adoption of BOT Technologies ➢ REDUCE Cost of Adoption ➢ MAINTAIN Ongoing BOT Services ➢ ENABLE Conversational Commerce ➢ BUILD AI-enhanced Business Decision Body of Knowledge ➢ USE ANY BOT Tech: Slack, FB, Twitter, IRC, YouTube, Reddit... ➢ GENERATE Fees from Transaction Servicing
- 26. IT’s STILL EARLY DAYS: Step 2
- 27. Questions? The Blueberry Muffintops Hackathon Team