Subscribe
Sign in
Init.ai — Build powerful and intelligent conversational apps 💁🤖
Create AI and natural language conversational apps through our developer platform. We offer a comprehensive solution that handles messaging, machine learning, and business logic, including integration with third-party APIs. You can suggest responses and update agent workflows in real time as well as automate many conversations.
46
Replies
@gkoberger thanks! Can't wait to see what you build.
BTW -- we love hosting our docs on ReadMe.io. Great product.
Such a smart and useful product for ecommerce and related fields. Excellent work Init.ai team!
@moloneymike Thanks Mike! We see conversation as transformative in e-commerce and many many other industries. We're starting with automation, but our approach enables us to work with multi-user conversations and assist sales and support staff as well, so the possibilities are really endless.
Super cool.
Do you see conversational apps as an additional channel for chatbots or the main one?
@harryraymond Hi Harry, thanks for the question!
We see chatbots as a subset of conversational interaction -- where the broader term can mean full automation (either natural language or with a more restricted button-based interaction), two or more people talking, or agents helping customers through live chat. We're targeting this broader space in general while enabling companies to build better, higher value bots.
These interactions are happening everywhere: messaging platforms like Facebook, Telegram, Kik, etc, embedded with websites and mobile apps, and increasingly through voice on IoT devices. We help companies deploy apps on every channel.
I've been working with @keithbrisson and @adawoodi as they took the first steps in building init.ai's messaging connectivity with Smooch. At every turn, I've been impressed by their technical prowess and the potential of their platform as one of the first "next-generation" conversational experience builders.
This tool may not be your best choice if you want to create a relatively simple bot experience based on keywords, however it's a great way to build, train and maintain a more sophisticated conversational experience - particularly if you are a developer that wants to shift into high gear and really control your outcome.
@keithbrisson - can you tell us more about the insights that are behind your unique approach? How does init make bot building better?
Hi @gozmike, thanks so much for the kind words and question.
We're providing developers the tools to make great conversation experiences that add real value to the companies they work for. We think we're just at the start of this great shift towards chat and natural language conversation, and we want to power it now and into the future.
To make that happen, we started with the basics. We built a unique, conversational NLP system that works with messages in sequences instead of isolation, to better understand context and meaning. In end-user trials, people always wanted to go beyond questions and answers to a back-and-forth interaction. That's what we ultimately want to enable.
And then we built all the tools that make training such as system possible. The machine learning underneath is sophisticated, and always changing as new techniques come out, but we abstract it away into simple developer tools to let companies focus on their core business, not arcane technology.
Finally, we see conversational interactions emerging everywhere--from messaging apps to wearables to IoT--so we enable these applications to be built once and deployed everywhere.
@leftearpod Hi Jake!
Quick answer: we tried building a conversational app on existing tools and none fully suited our needs!
Long answer: we see a huge shift taking place right now towards people connecting with each other and with businesses though natural language chat. For consumers, usage of messaging applications has been growing like crazy. For businesses, demand for live chat has been steadily increasing and is dominating other channels for consumer interaction.
We're also at this interesting inflection point where computers are starting to be able to join in the conversations. The possibilities there are really amazing -- especially around simplifying information retrieval as the world around us becomes more data-driven and complicated -- and just getting started.
We recognized these trends starting a few years ago, when we realized the incredible potential to simplify interacting with complex datasets. And we thought long and hard about what the user experience would need to look like to make that work. Ultimately our goal is to enable simple, powerful, robust interactions that simplify consumers' lives. It's a very UX-driven approach to starting a company that had ended up developer oriented and focused on relatively serious tech.
Drag & drop bot authoring tools are very cool, and can get a bot built quickly, but keyword-based bot triggers can be limiting, and a authoring a graphical flow becomes difficult with a complex bot script.
NLP can make a bot experience much smoother, but again, unless you invest massive amounts of time into your NLP model and training set, your bot is going to frustrate users. When NLP based flows go off the rails, it's also difficult to pinpoint what went wrong.
I played with Init.ai and was very impressed. init.ai offers a much better toolset for developers to build, train, and debug natural language buts, and I can't understate how important debugging is. Bots and NLP are an inexact science, so having a good debugging tool is crucial, and the one Init.ai has built is mind-blowingly awesome.
Congrats to @keithbrisson and the whole init.ai team.
Hello Product Hunters!
We’re quite psyched to launch today here. We’re Init.ai (Techstars Boulder 2016), a simple-to-use platform that makes building intelligent conversational apps easy.
We want companies and brands to go beyond the limits of chatbots. With our platform, they can engage their users naturally. Developers can build apps that would be up and running in less than hour.
You can start building your conversation apps right now — try it yourself, free!
Looks cool! Did your team build the NLP from scratch, or is it powered by something else?
@markcwebster Hi Mark! Yes, the NLP was built from the ground up and is internal to our platform.
Our strategy is to take the most recent advances from the machine learning field, combine and extend them in conversation-specific ways, and then abstract it all into a simple workflow to make it accessible to any company. For a few tasks we use open source tools (like parsing dates, etc) but all the core tech is based on various neural network types capable of handling ongoing conversation (and we're constantly evolving the implementation under the hood).
We want to let companies reap the benefits of new language understanding research as it emerges, making the developer experience simpler while improving the end user experience -- all without the company needing to know anything about machine learning itself.
@keithbrisson Init.ai looks awesome!
If you want to add some fun to your platform check out Guggy - https://guggy.com
This is great. I'm thinking about playing around with this but I'd like to at least have a ballpark for pricing; I don't want to spend time on it and then have to drop it because of price. It looks really good and it'd be a shame to not use this!
Hi @keithbrisson - This is very cool. Any ideas when the REST API will be available? I'm currently using LUIS but it doesn't hold context too well. Also, it's great that you're free whilst in beta but some idea of pricing is needed because I don't want to invest in this and discover it's costly later on. Thanks!
It's exciting, and overwhelming, to see the rate at which these tools are popping out.
One thing that really caught my attention about Init.ai was "Assist". I saw you were using Intercom on your site, which seems like a related use case.
Do you expect bots to start popping out outside of chat clients? Would you say "Assist" is an example of that?
I have been using init for about 1 month, and it works great! Slack is very active and everyone is there to help! I have had some suggestions and the team works on them right away. Works great with Smooch.io and Facebook as well!
The init.ai team is top notch! I had a great time working with them(and making awesome bots) during Techstars Boulder 2016
What would you say are the differences between Init.ai and Wit.ai? Why would one use Init over Wit?
@hickmanzackary Hey Zack, all products have their use cases. We built Init.ai to handle conversation flows, understand context (i.e. is “cool" a temperature or another word for “hip"), and integrate tightly with messaging, mobile, or web. We also have a strong developer workflow with versioning, integrated custom logic, etc. Plus soon you will be able to handle multi-user chats or even exchanges where bots monitor conversation and interject as needed. We think the Wit.ai team is great and hope you’ll try our product and let us know.
ReadMe