ProgrammableWeb's Most Interesting APIs in 2017: Cognitive Computing

This next section of interesting APIs from 2017 contains notable choices from our Artificial Intelligence (AI), Recognition, Machine Learning, Predictions, Augmented Reality, and Natural Language Processing categories, beginning with three offerings from Google plus three more from Facebook and Microsoft. Incidentally, our two most popular hashtags on Twitter (besides # API), were # AI and #MachineLearning, which must mean developers are quite interested in these categories of APIs!

Google's Cloud Video Intelligence API makes it possible to scan and search videos for nearly anything thanks to Machine Learning. Developers can use the Google Cloud Video Intelligence API to scan every single frame of videos for specific images. This API is driven by machine learning technology and can seek out key subjects and tag metadata found in a video.

Google Cloud Video Intelligence API recognizes and identifies objects in videos Video: Google Cloud Platform

Google's ARCore aims to bring augmented reality to Android handsets and other devices. The ARCore API can be accessed via SDKs for many of the most popular development environments. With the API, AR features like motion tracking, environmental understanding, and light estimation are accessible. ARCore works with Java/OpenGL and the Unity and Unreal engines. Read more about ARCore here.

View a preview of a new SDK called ARCore from Google Video: YouTube/Google VR

Google has recently introduced Dialogflow Enterprise Edition, its Platform for building voice and text conversational applications. Dialogflow was formerly known as API.AI. The Dialogflow Enterprise API (Beta) processes a natural language query and returns structured, actionable data as a result. This API is an end-to-end development suite for conversational interfaces including; chatbots, voice-powered apps and devices. A Dialogflow Enterprise gRPC API is also available. GRPC is the Remote Procedure Call system that was developed at Google.

Facebook Thoughts is a collaboration between Facebook and the neuroscience research community that has set a goal of typing thoughts straight from someone's brain. The Facebook Thoughts API provides programmatic access to thought-to-text services. Facebook Thoughts will only detect thoughts that people have sent to the speech centers of the brain, allowing users to choose what to share and what to keep private. Facebook is currently accepting applications for its beta developer program. This API is also referred to as the Brain API.

During October, 2017, Microsoft announced the general availability of two of its cognitive services APIs: the Bing Custom Search API and Bing Search API v7. Bing Custom Search API offers users access to targeted web search experiences and delivers an ad-free search engine. Examples of custom parameters include limiting searches to specific domains, subsites, and webpages. Bing Web Search API returns a list of relevant results from a search query which may include images, videos, news, related searches, and spell correction.

Spectrum provides intelligent algorithms for moderating content. The platform aims to prevent cyberbullying and promote thoughtful content. The Spectrum API can be integrated to flag content found on forums, chat channels, blogs, and other sources as inappropriate or toxic, and offers dashboard support with information such as user discussion index, and member retention.

Heat Intelligence is a customer service platform that helps companies help their customers through Artificial Intelligence response. The Heat Intelligence REST API allows developers to make requests and queries, retrieve answers based on parameters, embed videos, and make phone calls.

Heat Intelligence API

Diagram of how Heat Intelligence works Image:Heat Intelligence

IBM Watson Discovery provides functionality for searching, crawling, converting, and normalizing data and providing features for discovering patterns in data and insights from data. The IBM Watson Discovery REST API allows developers to build queries for data and retrieving results, add data, manage queries, and integrate with other IBM Watson APIs.

ConstructVR offers a platform and features to create and deploy virtual reality and augmented reality applications. The ConstructVR API allows developers to integrate the functionality of ConstructVR with other applications and to create new applications.

Gluon provides developers with a way to build machine learning models using a simple Python API and a range of pre-built, optimized neural network components. The Gluon API, a joint offering from Microsoft and Amazon, offers an interface that allows developers to prototype, build, and train deep learning models. The Gluon interface works with the deep learning Framework Apache MXNet and will support Microsoft Cognitive Toolkit (CNTK) in an upcoming release.

With, developers can kickstart the machine learning process by starting with rules that learn and improve automatically over time. API allows any developer to easily add machine learning to their working code. The API is useful for applications concerning product recommendations, content recommendations, dynamic pricing, matching in marketplaces and customer scoring into Web, mobile and conversational apps.


A Fuzzy agent is a virtual intelligent machine added to code for complex decision-making Image:

Deep Affects provides Emotional Intelligence (EI) technology for Human Resources and Project Management. The service aims to analyze and improve team performance and company culture. The Deep Affects API offers several tools used to extract analytical emotion data from audio and text. Developers can integrate emotional intelligence capabilities such as individual recognition, denoising, and para-linguistic features like pitch, rate, and amplitude.

Deep Affects dashboard

The Deep Affects dashboard Image:Deep Affects is a platform that helps users build and deploy intelligent conversational bots and applications with artificial intelligence. The REST API allows developers integrate the functionality of with other applicationsand manage users, manage messages, access conversations and more. uses AI to curate real-time financial regulatory information. has recently launched a new developer portal with an API, Documentation, code samples, and other tools developers can use to build applications integrated with financial regulation data. Sources include rules and enforcements from federal & state level agencies, executive orders, regulatory publications, press and independent agencies. The API retrieves data via a SMART platform (Search, Monitor, Access, Research and Track financial regulatory information). developer platform overview developer platform overview Image:

Q emotion's AI solution has been developed based on millions of tests on actual feedback. The Q emotion Text to Emotions Premium API detects emotional indicators in submitted text to determine the emotional index, key primary emotions, and main sensations expressed in the text. This API can be used to monitor customer experiences, build products, and improve text-base communications.

SnatchBot is an artificial intelligence messaging bot builder tool. The SnatchBot API enables bot creation for mobile devices, and web apps to send and receive messages. With the SnatchBot API, developers can implement sentiment, extract URL, include statement, add multiple choice, await response, extract emails, and collect numbers.

Twilio provides cloud services for building messaging, voice and video capabilities directly into applications. Twilio DataTrack API and Twilio Media Sync API (coming soon) are new APIs for enabling multi-user augmented reality (AR) applications. The APIs are part of Twilio's Programmable Video platform. The DataTrack API allows developers to create a DataTrack channel for sending low latency messages to zero or more receivers subscribed to the data. This allows for building collaboration features like whiteboarding, screen annotations, shared augmented reality apps and more.

WebSanitize allows users to filter out unwanted words from text on web and mobile applications and text. The WebSanitize REST API inspects text bodies for profanity and swear words, replacing with an alternative word or special character.

Matroid allows people to search media such as videos and images for objects, faces, and other items. The Matroid API lets registered account holders get programmatic access to the technology for third-party applications. The API is listed under the Recognition category. See ProgrammableWeb's complete list of Recognition APIs.

Matroid overview and demonstration Video: YouTube/Matroid

ParallelDots provides AI-driven visual analytics and text analytics APIs. Among their several API offerings is the ParallelDots Emotion Analysis API, which Detects emotion attached to textual data on social media, surveys and customer feedback. Also the ParallelDots Abusive Content Classifier API which detecs and filters abusive content present in any textual data. And the company recently released a Virality Detection API, which predicts anticipated popularity of a photo based on photos trending on social media.

Kite AI API enables developers to determine the overall mood of the conversation, or track conversation flow between messages in order to to solve online harassment issues. The technology uses bot AI and human moderation to determine offensive comments.

Kite AI says their API can analyze text in less than 150ms with 90% accuracy

Kite AI says their API can analyze text in less than 150ms with 90% accuracy Image: Kite AI

Voysis is an artificial intelligence (AI) voice platform that provides natural language speech processing for voice-controlled applications. Voysis REST API methods include retrieving conversations, retrieving text-to-speech, and managing conversations.

Stickers are similar to animated emojis intended for messaging services. The Stickeroid AI API is a cloud platform for stickers based on neural networks that translates text to sticker for most any situation of conversation. The API can replace about 250,000 words into stickers. Developers can use the API to query sticker type, id, full image or thumbnails, size of images, and more.

Stickeroid AI translation examples

Stickeroid AI translation examples Image: Stickeroid AI

Amazon Web Services (AWS) Lex API is built on the same deep learning technologies that Alexa is, and allows developers to create conversational interfaces and conversational bots.

What did we miss? Submit your Cognitive Computing API choices in the comments section or via Twitter.

Be sure to read the next Artificial Intelligence article: Daily API RoundUp: Salesforce Einstein, ThingScale, Mailforce