Overview

pSpoke is a modern ticketing system to manage workplace requests that uses Machine Learning to automatically answer questions and assign requests to right teams./p pSpoke goes beyond traditional ticketing with their friendly, AI-powered chatbot that gives workplace teams hours of time back as it automatically responds to questions on Slack, email, SMS, and web. And the more employees ask, the more Spoke learns./p p /p pStarted in August, 2016./p pBased on SF. Funded by Greylock and Accel./p pSmall, fast-moving engineering team./p

Supernova Award Category

Future of Work: Employee Experience

The Problem

pNatural language processing problems:/p pCustomized machine learning models for every client/p pNeed to learn quickly (near real time) from user interactions/p p1000s of ML models/p pOperationally, as the company grew, they had issues scaling within the stringent tiers and a small dev-ops team with many responsibilities. After a particular issue involving a 5-megabyte query that took them days to debug, the team started to look for another service where they could have more visibility and prevent issues before they happen./p

The Solution

pFast Online Machine Learning with MongoDB/p p /p pOnline learning: Update the model at each time step as the data sequentially arrives/p pSpoke built quickly by using online learning to deliver a slick product experience. Users see the utility because the system learns in real time!/p pEasy to serve and scale using MongoDB/p p /p pPerformance reliability and powerful database management tooling with MongoDB Atlas/p pSignificant increase in productivity, with more time to focus on developing core features rather than performing database operations and management tasks/p pWent from tracking just 3 metrics to nearly 30 metrics, including CPU load and task processing times /p pWith Performance Advisor’s integrated alerts and the optimization suggestions, the team can now anticipate and prevent issues before they arise. /p

The results

pMongoDB’s flexible document model, dynamic indexing, scalable atomic operations, powerful database management tools, and performance reliability enable Spoke to seamlessly embed artificial intelligence in every aspect of its workplace productivity solution. Spoke’s modern disruptive ticketing system, awesome Slack integration, and self-service A.I. save companies time and makes employees happier./p

Metrics

pSpoke deploys scale to 1000s of machine learning models and achieves sub 500 ms fetch enabling fast online learning that delivers immediate results. When users ask questions to Spoke they get accurate real time responses selected from 1000s of potential answers using a combination of MongoDB test search, text indexes and machine learning./p p /p pOperationally, poke went from tracking just 3 metrics to nearly 30 metrics, including CPU load and task processing times. With Performance Advisor’s integrated alerts and the optimization suggestions, the team can now anticipate and prevent issues before they arise./p p /p p“We used to spend eight engineering hours once a week or so trying to figure out what was going on, and that doesn't exist anymore.”/p p-Roopak Venkatakrishnan, Software Engineer, Spoke/p

The Technology

pMongoDB Atlas/p pMongoDB Driver for Python (PyMongo)/p pAtlas Live Migration Service/p

Disruptive Factor

pUsually products that utilize machine learning or AI acquire and prepare lots of data, train and tune models, then deploy and possibly update them periodically delaying time to market and user adoption. Spoke is different because it continually adapts to user actions in an accurate and agile way using real time data and interactions, increasing employee engagement and improving productivity. Initially, Spoke encountered issues while scaling with another managed service provider, but after making the move to MongoDB Atlas (which took just fifteen minutes using the Atlas Live Migration Service) the team has seen a significant increase in productivity, giving them more time to focus on developing Spoke’s core features rather than performing database operations and management tasks. /p

Shining Moment

pMongoDB and its intelligent data platform architecture, which includes flexible document model, and powerful and dynamic indexing capabilities, also helps Spoke with tasks such as text searches, session management, and fast retrieval of callback maps to enable conversational dialog. These tools help ensure that Spoke and its AI infrastructure can learn and adapt to client needs at scale while revolutionizing workplace productivity./p