What Technical Solutions Are Used in AI?

DZone 's Guide to

What Technical Solutions Are Used in AI?

Here's what 22 executives who're familiar with AI and all its variants said when we asked them, "What are the technical solutions you/your clients use for AI initiatives?"

· AI Zone ·
Free Resource

To gather insights on the state of artificial intelligence (AI) and all its variants — machine learning (ML), deep learning (DL), natural language processing (NLP), predictive analytics, and multiple neural networks — we spoke with 22 executives who are familiar with AI.

We asked them, "What are the technical solutions you or your clients use for your AI initiatives?"

Here's what they told us.

Data Science

  • Apache Hadoop, Spark, Apache Drill.
  • The most prevalent is TensorFlow, followed by Café and Torch and other database platforms. We enable these platforms to pick the right training sets with billions of options. Feature-rich data sets are able to differentiate between attributes and applying the appropriate analysis. Provide a training feature set to train on. Inferencing is involved onto high-velocity machines accelerating time to production.
  • We use a combination of algorithms with deep learning for anomaly detection. You need to think about what algorithm to implement to solve a particular problem. This ranges from machine learning to deep learning algorithms.
  • We build an abstraction layer on top of TensorFlow providing a set of tools and a platform that go from the mechanics of learning to subject matter expertise and domain expertise. We see everything. A client using Watson and NLP to experiment directly. Organizations are exploring what they can do with AI. Take stock of what’s out there. What are the capabilities to form strategies for tackling? How to tackle real-world control and optimization problems? We provide a higher level of languages and tools to help accomplish a higher level of detail.
  • The context is different for different people. We use a lot of ML algorithms — TensorFlow, NVidia, modified TensorFlow GPU, Intel Titan.
  • Python and R are hot now. MatLab is the old classic. More data science start-ups are building platforms that allow for quicker analysis.
  • We use Spark instead of Hadoop for many AI initiatives. It’s easier to deploy and scale.
  • We embrace open source to accelerate people getting started — R, Python, H2O, TensorFlow, MatLab — very heterogeneous.
  • Apache Spark, machine learning algorithms, deep learning with TensorFlow. Neural network standard libraries. Python with the Panda library. Open source tools people can use. Tools from Google and Facebook. Use APIs to access. There are a lot of different open source tools.
  • Open source like Apache Spark embedded for AI algorithms. TensorFlow is our AI platform for deep learning. A lot of different complementary technologies for chatbots. All cognitive services like Watson Dev Cloud, Azure, and Google. We leave our platform open so developers can take advantage of APIs.
  • Spark, TensorFlow, Google open source, Microsoft libraries, and Kafka are changing how we code, build algorithms, massage data, and wrangle data.
  • Creating solutions composed of a set of services. A range of languages: Java, JavaScript, Python, Go for edge and mobile. Off the shelf services for NLP, discovery, speech detection through APIs. When building your own algorithms to solve quantitative data problems Python, R, and IDEs are surfacing like Darviz and Node Red for wiring components.
  • R and Python. Used to see a lot of SAS but not anymore. Open source tools. ML as a service via AWS, Azure, and Google Cloud.

Other Tools and Platforms

  • There are many players like IBM Watson, Microsoft, Google, and start-ups giving developers access to AI through APIs. API.ai is a service for chatbots bought by Google. There are many different libraries available. Use one of those and write your own code realizing different platforms have different strengths – IBM for NLP, one for virtual bots. We write our own code for machine learning (ML). There are new technologies and the need for new skill sets. How do you know who has the best chatbot platform? Should you put your bets on a big platform or a start-up?
  • Most ML capabilities form standard AI, linear regression, decision tree models. Areas around data linkage with domain specific algorithms that will build complex social graphs. Collusion is an important component of fraud. You must look at social networks for healthcare fraud, law enforcement, financial fraud, and tax fraud.
  • We are actively exploring ways we can integrate intelligent chatbots and voice assistants into our business. We have implemented Alexa functionality and are experimenting with Google Home and Siri.
  • Our technology works in a variety of ways but one of the most popular solutions is due diligence exercises. For example, organizations would historically have to manually provide due diligence checks and recommendations to thousands of their contracts. This process can be extremely laborious and firms end up having to extend working hours or draft in supplemental resources. Our technology can automate this process, freeing up resource and increasing efficiencies.
  • We have another solution for contract analysis where the technology can automate the clustering of contracts and then extract key data points from each category. We recently created specific business solutions that help organizations comply with regulatory changes. For example, the GDPR solution can help facilitate GDPR compliance by automatically identifying documents and other types of data in any business system which is subject to GDPR rules, allows users to view feeds on the latest personal data that requires attention as well as expedite requests for information. We also recently launched an analytical tool to predict and forecast outcomes by analyzing historical data. A typical use case for this is the ability to predict the cost of legal matters and other types of projects using AI algorithms on data that has been surfaced using our platform.
  • If you think about how to apply predictive analytics to help make decisions, it frequently leads to better outcomes and also saves on costs. We’ve worked with hospital networks to use information about patient trends, such as how many people were coming into the hospital, how many beds were full, how many beds were expected to be full a week or a month from now, to inform intelligent staffing for nurses and doctors. We pull from 18 different data sources to create predictive data models. In this instance, AI was used to tell a hospital that based on what’s happening today and what’s expected to happen in the next few weeks, here’s how you should be staffing the different areas. That kind of knowledge has the ability to positively impact the quality of care transforming the patient experience in addition to helping the workforce. AI is also being used to make recommendations for treatment that will help reduce the lengths of stay, reduce labor costs and improve health outcomes. It’s a win-win and the holy grail of healthcare patient service and workforce enablement.
  • We’re leveraging deep learning technology. Unlike machine learning, we train the deep learning neural network, similar to a human brain, to learn, adjust and respond to threats. The deep learning offering can detect malicious behavior across multiple vectors, and provides true adaptive defenses against the most advanced cyber attacks. We are the only company providing endpoint protection platform (EPP), mobile, and remediation capabilities.

What are the technical solutions you see being used in AI?

Here’s who we talked to:

  • Gaurav Banga, CEO, and Dr. Vinay Sridhara, CTO, Balbix
  • Abhinav Sharma, Digital Servicing Group Lead, Barclaycard US
  • Pedro Arellano, VP Product Strategy, Birst
  • Matt Jackson, VP and National General Manager, BlueMetal
  • Mark Hammond, CEO, Bonsai
  • Ashok Reddy, General Manager, Mainframe, CA Technologies
  • Sundeep Sanghavi, Co-founder and CEO, DataRPM, a Progress Company
  • Eli David, Co-Founder and Chief Technology Officer, Deep Instinct
  • Ali Din, GM and CMO, and Mark Millar, Director of Research and Development, dinCloud
  • Sastry Malladi, CTO, FogHorn Systems
  • Flavio Villanustre, VP Technology LexisNexis Risk Solutions, HPCC Systems
  • Rob High, CTO Watson, IBM
  • Jan Van Hoecke, CTO, iManage
  • Eldar Sadikov, CEO and Co-founder, Jetlore
  • Amit Vij, CEO and Co-Founder, Kinetica
  • Ted Dunning, PhD., Chief Application Architect, MapR
  • Bob Friday, CTO and Co-founder, and Jeff Aaron, VP of Marketing, Mist
  • Sri Ramanathan, Group VP AI Bots and Mobile, Oracle
  • Scott Parker, Senior Product Marketing Manager, Sinequa
  • Michael O’Connell, Chief Analytics Officer, TIBCO
ai, big data analytics, machine learning

Opinions expressed by DZone contributors are their own.

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}