美国创业孵化器Y Combinator孵化的13个机器学习和数据科学初创企业简介与启示

中国IDC圈如果你正在4月21日报道，筹划开发一个产品，建议优先考虑一下这几个初创企业。也许就会发现一个产品新视角，那就是通过机器学习和预测分析法使产品性能更强大。

这些初创企业是在2016年冬季Y Combinator 公司中脱颖而出的。Y Combinator 是一个初创企业孵化器，每年两次将$120k 投资于初创企业。像Reddit, Quora, 人工智能rbnb, Dropbox这些的成功企业都是因Y Combinator而走入公众视野的。

“帮助企业”的理念正是这些初创企业的核心。这些人能敏捷地洞悉发展趋势、灵活地驾驭科技使商业更加简单快捷。对于你过去可能需要用两天完成的工作，现在的一些初创企业几个钟头就可以帮你搞定。

以下所列的一些初创企业看起来非常离经叛道， 可又让人觉得前景光明。人们通过深度学习和人工智能设计产品的方法一定会让你大开眼界。

注意：本文是对从现在到未来几年把自己定位为数据科学企业家、并渴望从事于商业想法的读者的启发性文章。

你将收获什么？为什么要关注？

我们很多人都担心机器学习和人工智能的未来。有个叫Quora的流行论坛里充满了诸如“机器学习在2020年发展到什么程度’一类的问题。如果你对于数据科学的未来也存有类似的疑问，那么你将在这里找到满意的答案。

近来人们对机器学习和预测分析学的认可飚升。几乎各行各业（石油和金属业我不确信）都将数百万投资于执行数据驱动的商业方法。这些初创企业将定义数据科学业的未来，并成为扩大就业和增加商业项目的主力军，而它们其实仅仅是很小的一部分。还有很多企业没有列在名单上，因此一定要相信这个行业的未来。

通过本文你会了解到机器学习和数据科学里的前沿商业观点。也许明年站在Y Combinator台上分享商业妙点的就是你！

机器学习/数据科学初创企业

Msg.ai

大多数人都喜欢发简讯。于是Whatsapp, Facebook, Snapchat这样的APP在前期就吸引了上百万的融资，考虑到即时通讯这一媒介迅速走红，Msg.人工智能 公司就提供了这样一款运用机器学习和深度学习与客户进行智能互动的即时通讯软件。这个软件可以使公司在所有的热门即时通讯平台上做宣传并获得所有的热门手机应用APP。此外，该产品还能进行情感分析、趋势分析，并提供详细的报告。

Protonet

一个公司凭什么不能使用公有云服务器？这个初创企业将给你所有答案。他们就推出了一种更安全的通过私有云服务器来共享数据的方法。Protonet是针对安全私有云服务器的一种项目管理和协作软件来帮助企业实现数据安全。该产品的开发是要满足高运行、更快数据接入和上传的需求。其多种特点包括任务管理、文件共享、商业沟通和移动协作等。

Interstate

营销手段已经变得越来越智能。各个公司再也不会盲目投入，而是理解了营销分析学的重要。Interstate就是一个免费的营销分析和归因平台，它运用预测建模帮助企业优化预算分配、锁定正确的客户群、帮助客户省钱。它还提供动态的营销支出数据，实现仪器的十字追踪并与世界热门APP的对接。

Elucify

生成前导数据很耗时。大量时间浪费在了应付数据。Elucify运用人工智能和机器学习帮助企业从前导数据中获得最多效益。该产品使公司能辨别和清理陈旧的引头，提取目标引头进行快速转换，动态地搜寻和更新前导数据，并在几个小时内完成这一切。说到底就是，速度至上！

GitPrime

GitPrime得名还是源于热门的GitHub社区。GitPrime的目标就是使软件工程更顺畅更灵活。它因使用统计分析、辨别模式使工程初学者掌握构建软件的方法，提供个性化的建议而闻名。此外，GitPrime使软件工程团队能沟通、共享和跟踪团队进展等。GitPrime致力于通过提供基于数据的时时反馈和分析运行情况使软件工程师更加高效。

Nova

Nova的目标是通过使用数据科学的方法来提高销售团队的有效性。这一点通过一个即时通讯平台即可实现。该平台使代表根据最佳回复发送定制化的电子邮件。该软件使用文本挖掘，分析情感，针对具体的人群发送最适合的邮件。通过软件，公司称发送能力（减少垃圾邮件）已经提高，3倍于之前的速度，回应率也有所上升。

Enflux

如今人们比过去更加关注健康。他们意识到健康比财富更重要，因此尝试接纳任何可能的事情和习惯。该初创企业提供内置传感器的机洗衣服，它可捕捉一个人完整的活动动态。简单说来，这个运动衣可以记录身体运动并生成数据。一旦获取了数据就能传达特定的身体信息。谁会想到我们穿的衣服有一天都能记录数据呢？

Hykso

拳击手能使用分析学吗？即使他们不能，Hysko也将这变成了极具影响力的现实。这个初创企业已经推出了一个产品，按照速度、强度、计数记录每一个动作，并生成分析结果。笔者发现这与fitbit app很类似，不同就是专为拳击手而设置的。Hysko提供实时检测、审查运行结果，帮助拳击手快速克服障碍。不用说，这种装置也是内嵌传感器来收集数据并转化为数字。

Skymind

日益增长的大容量、多种类数据正在扩大机器学习的需求。科学技术飞速发展，其中Skymind就是一个强劲的后起之秀。Skymind的目标就是运用可量化的深度学习为企业提供先进的分析解决方案。这一软件配有Hadoop, Spark, Neural nets和其他重要的部分。目前他们的工作主要是在金融和营销行业。

Wakie

想过让一个陌生人叫你起床吗？Wakie app的功能就是这个。现在，你可以和“友好”的陌生人讨论任何话题、建议而不用交换联系方式。从深层次上讲，他们使用机器学习分析用户的知识和技能。最终，选择最佳用户回答问题。此外，他们还使用语音识别技术来评估所提供答案的质量。

Zenysis Technologies

集中数据生成结论绝非易事。Zenysis技术已建立了一个数据集成软件使编纂和分析同时进行。可以说这就是一个数据分析软件。通过这个，他们计划使用其大量生成的数据帮助政府和发展中国家提高管理。他们提供可行的观点帮助国家进行有根据的决策，消除决策中的任何模棱两可因素。

PaveIQ

不确定投资是否获得了高的收益？你的营销策略带给你应有的回报了吗？像Pave IQ这样的初创企业现在来救你啦！Pave IQ是一个营销分析公司，帮助企业使其运用谷歌分析这一网站流量统计与分析工具得出的观点更加可行。它分析公司的目标，通过各个渠道使用机器智能来提取营销结果。它同时以数据为基础，提供关于如何提高投资回报（ROI）的定制化报告。

DeepGram

从视频中提取观点和情感？现在这也实现啦。DeepGram研发了一款工具，建立人工智能模型自动分析和归类视频/音频流。它无需重复暂停视频即可锁定所需的关键词。此外，它还使用深度学习算法，提取话语为文本观点，免除了手动过程。这在呼叫中心、视频数据组和媒体中心都非常有用。

尾注

以上的初创企业都具有独特的价值。即使这些产品已然问世，不过产业规模依然巨大，而富有创新的人们总能给社会带来积极的影响。在120个初创企业中，13个公司已研发了基于机器学习和人工智能的产品。这表明，世界各地的人们不仅会关注人工成本，还会考虑完成工作的速度。

在本文中，我从2016年冬季Y Combinator 公司孵化的项目中列举了13 个机器学习和数据科学领域的初创企业。正是他们的创意值得你的留意。这有助于帮你理解多种应用和技术的使用。不要认为机器学习或分析学局限于某个行业。相反，各行各业的人都已经开始寻找使用它的方法了。

原文：

13 Machine Learning & Data Science Startups from Y Combinator Winter 2016

Introduction

If you’ve been planning to build a product, I’d suggest you to check these startups first. May be, you can find a new angle to your product and make it more powerful using machine learning & predictive analytics.

These startups got featured at Y Combinator Winter 2016. Y Combinator is a startup accelarator which invests ~ $120k in startups twice a year. Successful companies like Reddit, Quora, Airbnb, Dropbox are known to emerge from Y Combinator.

The essence of ‘helping businesses’ lies in the heart of these startups. These people are smart at reading trends, harnessing technology to making business easier and faster. If you were taking 2 days to finish a job, some of the startups might help you finish it in hours time.

You’d be amazed to see the kind of idea running through peoples’ brain around the world. Some of the startups listed below are simply incredible and looks to have a promising future. You’ll see how people have designed their products using deep learning and artificial intelligence.

Note: This is an inspirational post for people who see themselves as data science entrepreneurs in few years from now and aspire to work on a business idea.

What will you learn and Why should you care?

Many of us, worry about the future of machine learning & artificial intelligence. A popular forum Quora is filled with questions like ‘How far will machine learning reach in 2020?’. If you too have similar question pertaining to future of data science in mind, here you’ll find a good reason to trust.

The acceptance of machine learning and predictive analytics has seen an overwhelming surge. Almost every industry (I’m not sure about oil and metals), has invested millions of dollars in implementing data driven business methods. These startups will define the future of data science industry. Such startups will be the main source of increased jobs and business projects. And, these are just the cherry picked ones. There are many more which couldn’t become a part of this list. Hence, be assured about the future of this industry.

With this post you’ll learn about upcoming business ideas in machine learning and data science industry. Who knows, you might end up presenting your business idea at Y Combinator next year!

Machine Learning / Data Science Startups

Msg.ai

Most of us love doing messaging. As a result, apps like Whatsapp, Facebook, Snapchat have gathered millions of users worldwide. Considering the rising popularity of messaging medium, Msg.ai provides a messaging software which uses machine learning & deep learning to intelligently interact with customers. This software allows companies to establish their presence in all popular messaging platforms and accessing every app from this software. Also, this product is capable of performing sentimental analysis, trend analysis and provide detailed reports.

 Protonet

This start up has all the reasons as to why shouldn’t a company should not use public cloud servers. As a result, they have introduced a more secure way of sharing data using Private Cloud Servers. Protonet help companies to fight against data security using a project management and collaboration software on secure private cloud servers. The product is built to deliver high performance, faster data access and uploads. It’s multitude of features include task management, file sharing, business communication, mobile collaboration and much more.

Interstate

The methods of marketing have become much more intelligent. Companies no longer spend recklessly but have understood the importance of marketing analytics. Interstate is a free marketing analytics and attribution platform. It applies predictive modeling to help companies make better budget allocation, target the right set of customers and help save money. In addition, it also provides dynamic marketing expenditure data, allows cross device tracking and integration with world’s popular apps.

Elucify

Generating lead data is time consuming. Majority of the time gets wasted in data wrangling. But, with AI powered solutions Elucify aims to speed up and strength the complete sales conversion process. Elucify helps companies to reap the most benefits out of lead data using artificial intelligence and machine learning. Their product allows companies to identify & clean old leads, extracts the targeted leads for fast conversion, dynamically search and update lead data and do all this in couple of hours. After all, speed matters!

GitPrime

GitPrime get its name from the popular community GitHub. GitPrime aims to make software engineering less painful and more scale. It makes use of statistical analysis and identify patterns to let enginners improve the method of building software. It is known for providing personalized recommendations. Also, it allows the software engineering team to communicate, share, tracks team progress with data and much more. It endeavours to make software engineers more productive by providing data based timely feedback and analyzing weekly performance.

Nova

Nova aims to enhance sales team effectiveness using a data science approach. This is accomplished using a messaging platform. This platform enables a representative to send customized emails according to the best suited reply. This software uses text mining, analyzes the sentiments and decides for a best suited email for a particular personnel. Using this software, company claims to have improve deliverability (less spam), 3X speed and increases shoots up representative response rate.

Enflux

People nowadays are more health-conscious than they used to be. They understand that good health is above wealth, and hence try to adopt every best possible thing or habit available. This startup provides machine washable clothes which is embedded with sensors to capture a person’s complete workout motion. In simple words, this workout dress is enabled to track body movements and generate data. Once the data is captured, it delivers specific body insights. Who wondered even the clothes we wear will some day track data ?

Hykso

Can boxers use analytics ? Even if they couldn’t, Hysko has made this possible with a large impact. This startup has come up with a product which tracks every move in terms of speed, intensity, count and delivers analytical insights. I find it similar to fitbit app but for boxers. It provides real time monitoring, reviews performance and helps a player to overcome hurdles quickly. Needless to say, it is fitted with sensors which helps to collect data and convert them to numbers.

Skymind

Growing volumes and varieties of available data is increasing the demand of Machine learning. Technology is growing fast and among them Skymind is an emerging but powerful player. Skymind aims to provide advanced analytical solution to businesses using scalable deep learning. Their software is well equipped with Hadoop, Spark, Neural nets and other important integrals. Looking at their current setup, their focus lies in Finance & marketing industry.

Wakie

Ever thought of asking a stranger to wake you up in the morning ? Wakie app has made this possible. Now, you can talk to “friendly” strangers about any topic, advice, suggestions without having exchanged your contact details. At a deeper level, they use machine learning to analyze the content with knowledge and skills of users. Finally, selects the best user to answer a question. Also, they use voice recognition technique to evaluate the quality of answers provided.

Zenysis Technologies

Putting data together to draw a conclusion is not an easy task. Zenysis Technologies has build a data integration software which will compile and analyse all the data at one place. A data analysis software one can say! With this, they plan to help government and developing countries to improve their administration using their huge sets of generated data. They provide actionable insights which will help countries to make informed decision making and remove any ambiguity from decision making process.

PaveIQ

Not sure if you are getting the best return on investments? Are your marketing strategies giving you the required output? Startups like Pave IQ have come to your rescue. Pave IQ is a marketing analytics startup which help companies to make their google analytics insights more actionable. It analyses a company’s goals and uses machine intelligence to extract the results of marketing through all channels. It also provides customized data driven reports on how to increase ROI.

DeepGram

How about extracting insights, sentiments from an audio ? Yes, this is now possible. DeepGram has created an AI enabled tool which builds AI models to automatically analyze and classify the audio/ video streams. It is known to locate the required keyword without needing to stop/pause the audio repeatedly. Also, it uses deep learning algorithms to extract speech to text insights relieving humans of manual process. It is highly useful in call centers, search from audio data sets and media centers.

End Notes

The startups listed above have a unique value to deliver. Even if the product exists, the industry size is huge that these people can bring a positive change to the society. Out of 120 startups, 13 startups are found to have built product empowered with machine learning and artificial intelligence. This suggests, people around the world are not only concerned about human efficiency but the pace of completing work!

In this post, I’ve listed the 13 machine learning & data science startups from Y Combinator Winter 2016 batch. You should check these out, mainly because of the idea. This should help you understand about the diverse applications and usage of these techniques. Don’t think machine learning or analytics is limited to a particular industry. Instead, people have already started finding ways of using it across all industries.