双语：8家惊人大数据新创企业——他们的商业模式或许值得你借鉴

大数据产业每周都在发生变化，除非你非常密切地关注那些处于新兴大数据创新者漩涡当中的人，否则一些新贵可能已经超出你的注意范围了。在不久的将来，我们希望听到更多关于他们的事情，以及他们正在开发的技术、服务和应用程序。当然，在他们试图游进的快速的浪潮之中，其他人可能已经沉没得无影无踪！

Crowdflower

Crowdflower自称是“数据丰富的平台”，在搜索他们要找的见解（方法）的时候，其使用众包的力量来筛选其客户的数据集。

来Kaggle的500万数据科学家的以类似的方式进行网络工作。然而，而不是专注于为Netflix或者google这样巨头解决一次性和大的问题，他们可以投入到工作中为各种规模的企业解决更多的日常数据难题。

Flatiron

Flatiron 正在用OncologyCloud帮助对抗癌症。他们的目标是要建立和癌症有关的医疗信息的最大数据库，并使其通过云向医生、患者、教师和研究人员提供。

它的工作原理是，目前只从癌症患者的一个非常小的样本收集数据 ，即有4％的患者参与注册临床试验。对病因和治疗方法持有关键线索的那些96%的数据财富往往仍然锁定在医疗记录和医生的笔记之中。

Flatiron 希望通过它的数据采集技术汇总这些信息，那么世界各地的专业人士都会有一个强大的数据驱动的工具，这将产生新的治疗方法和更高的存活率。

LendUp

LendUp已经建立了自己作为一个“负责任”的替代发薪日贷款公司的形象， 这种公司传统上以过高的利率收取短期贷款的，支持高收费，并且如果客户没有付款会配合以积极的收债行为。

他们的信用得分是建立在关于贷款的大数据分析之上，这意味着他们可以更准确地评估每一个应用程序的风险，并提供了更公平的利率。它们为个人提供了一个合法实现短期信贷利率相当于一张信用卡或银行贷款的途径。

Infobright

Infobright是一个“物联网数据分析平台”。

它们的算法和数据库体系结构被设计成帮助那些正在产生大量的机器驱动的数据的企业 ，比方说来自于在工业环境中安装在生产设备上的传感器中的数据。这些数据集可能会变得非常大，非常迅速，这项服务的目的是保持它的可控性。

Feedzai

进入市场后，各种规模的公司是大数据中发现最流行的用途之一便是检测欺诈，特别是如果他们依靠网上交易的收入。

Feedzai声称它可以降低贵公司的欺诈性交易处理的80%，如果它的预测算法发现风险太大的交易符合一个概要条件，那么它就会阻止这个交易。

全面减少欺诈手段将增加零售商的利润，这可以以更低的价格转嫁给消费者,希望这会给我们所有人省钱!

TAMR

数据管理是管理进入公司数据流，并确保正确的数据被收集以匹配手头的工作。TAMR发展的服务将自动识别、清洗和给分析工作准备必要的信息。

其前提之一是数据科学家需要一个他们正在研究的数据的清晰,简明的表示,这样才能得出可靠的结论,所以Tamr提供一个接口,把你所有的数据源连接在一起。 它也被设计通过机器学习算法来处理高度自动化的工作,如制定许多决策包括需要哪些数据、如何展示这些数据等。但是，当它做得决定在自己的薪酬等级之上时，它会“伸手”向人类专家团队中的最合适的人寻求帮助。

Appuri

客户流失是一个导致企业失败的典型原因。所以Appuri让企业用数据算法来跟踪整个客户生命周期 即每个客户从他们开始购买你的产品或使用你的服务的那一刻直到他们停止使用之间的接触和互动。

我们的想法是，这将让你深刻地知道为什么客户停止使用你的服务，这意味着你可以进行调整以阻止它的发生。如果你知道一个可以预期导致高流失率的积聚事件正在发生，你就可以发动反击运动，以赢回快要离开的顾客。

GNIP

Gnip旨在帮助企业理解,在社交网络中产生的、日益增长的数据山脉，并获得有益的东西。

公众也越来越多的乐意在推特，脸谱，Instagram的和许多其他平台上分享他们的生活细节（比如他们买什么）。这给分析自己的网络提供了一种方式，并希望在帮助他们解决问题或填补他们的需求方面获得宝贵的见解如何。

英语原文：

The Big Data industry is changing weekly and unless you pay very close attention to who’s who in the maelstrom of emerging big data innovators, these upstarts might have slipped under your radar. Expect to hear more about some of them, and the technologies, services and applications they are developing, in the near future. Of course in the fast currents they are trying to swim in, others may sink without a trace!

Crowdflower

Crowdflower calls itself a “data enrichment platform” and uses the power of crowdsourcing to sift through its customer’s datasets in search of the insights they are looking for.

Their network of 5 million data scientists work in a similar way to Kaggle. However instead of concentrating on solving big, one-off problems for giants like Netflix or Google, they can be put to work solving more everyday data dilemmas for companies of any size.

Flatiron

Flatiron is helping the fight against cancer with its OncologyCloud, They are aiming to build the biggest database of medical information relating to cancer and make it available to doctors, patients, teachers and researchers through the cloud.

It works on the principle that aggregated data is currently only collected from a very small sample of cancer patients – the 4% which take part in registered clinical trials. The wealth of data from the other 96% which could hold vital clues to causes and treatments often remains locked away in medical records and doctors’ notes.

Flatiron hopes by aggregating this information through its data capture technology then professionals around the world will have a powerful, data-driven tool at their disposal which will lead to new treatments and greater survival rates.

LendUp

LendUp has built a reputation for itself as a “responsible” alternative to payday loans companies, which traditionally charge exorbitant interest rates for short-term loans, backed up with high charges and aggressive debt collection practices if customers miss payments.

Their credit-scoring is built around a big data analysis of the lending landscape, meaning they can more accurately assess the risk of every application, and offer fairer rates of interest.  They offer a way for individuals with legitimate need to access short-term credit at interest rates comparable to a credit card or bank loan.

Infobright

Infobright styles itself as a “data analytics platform for the internet of things”.

Their algorithms and database architecture are designed to help companies who are generating large amounts of machine-driven data – for example from sensors attached to manufacturing equipment in an industrial environment. These datasets can grow very large, very quickly and this service is designed to keep it all manageable.

Feedzai

After marketing, detecting fraud is one of the most popular uses that companies of all sizes are finding for big data, particularly if they rely on online transactions for their revenue.

Feedzai claims it can reduce the fraudulent transactions that your company handles by up to 80%, blocking them if its predictive algorithms find that the transaction fits a profile which is too risky.

Reducing fraud across the board means increases in profits for retailers which can be passed on to consumers in the form of lower prices, hopefully saving us all a bit of money!

Tamr

Data curation is about managing the data flow coming into a company and ensuring that the right data is being collected to match the job at hand. Tamr has developed services to automate the identifying, cleaning and preparing the necessary information for analysis.

Part of their premise is that data scientists need a clear, concise representation of the data they are studying in order to be able to draw reliable conclusions, so Tamr offers an interface for drawing all of your data sources together. It is also designed to work with a high level of automation – making many of the decisions about what data to include, and how to present it, automatically through machine learning algorithms. However, it will “reach out” to human experts on your team – asking the most appropriate person for help when it needs to make a decision above its pay grade.

Appuri

Losing customers (churn) is a good way for a business to fail. So Appuri lets a business put data algorithms to work to track the whole customer life cycle – the contact and interaction each customer has from the moment they start buying your products or using your services, to the moment they stop.

The idea is that this will throw up insights into why your customers stop using your service, meaning you can put changes in place to stop it happening. If you know a build-up of events is taking place that can be expected to lead to high churn rates, you can launch a counter-attack, with a campaign designed to win back departing customers.

Gnip

Gnip is designed to help businesses make sense, and gain insights from, the ever-growing mountains of data being broadcast over social networks.

The public is getting more and more comfortable with the idea of sharing details of their lives (such as what they buy) over Twitter, Facebook, Instagram and many other services. This offers a way to analyze your own network and hopefully gain valuable insights into how you can help them solve their problems or fill their needs.