AI/Machine Learning/NLP/AR/VR 2018-03-13T16:28:30+00:00
Quantilus AI NLP ML

Artificial Intelligence, Machine Learning, Natural Language Processing, Augmented and Virtual Reality

Intelligent Automation for Work and Life

At Quantilus, we have been working with AI before it became cool (and scary). Our first foray was in the field of Natural Language Processing – which we used for automated grammar and style checks of written content. Subsequently, we built tools to classify untagged content in intelligent, usable ways, and to present it for consumption with a high degree of personalization. More recently, we have been working on personality assessment of individuals based on 1) the words they speak (a relatively simple task), and 2) changes in their facial patterns based on verbal and visual cues (a much more complex task).

Want to build Virtual Reality or Augmented Reality apps for your business? We built some of the first business-focused AR apps for mobile and wearable platforms through our SAP partnership. Our apps help technical support personnel visualize product models, and also let customers visualize retail products in empty space. With the added complexity of tight integration with backend ERP systems.

FEATURED WORK

Appliqant - Automated Video Interviews
APPLIQANT – the Automated Interview Robot. Disrupt recruitment through the automated screening of job candidates.
Appliqant - Automated Video Interviews
Wearable Apps for Technical Support

Discover Simple Assist – Wearable Apps for Technical Support

Wearable Apps for Technical Support
Visual Showroom - Augmented Reality

Visual Showroom – Augmented Reality App for Product Display

Visual Showroom - Augmented Reality
Deloitte - Automated Document Editing, Data Extraction

Deloitte – Automated Document Editing, Validation and Data Extraction

Deloitte - Automated Document Editing, Data Extraction
Intel - Machine Learning

Intel – Machine Learning for Customer Classification and Segmentation

Intel - Machine Learning
Intelligent Lease - Data Extraction using NLP

Intelligent Lease – Automated Data Extraction from Unstructured Lease Documents

Intelligent Lease - Data Extraction using NLP
BluePencil - Grammar and Writing Style Checks using NLP
BluePencil – Automated Grammar and Writing Style Checking using NLP
BluePencil - Grammar and Writing Style Checks using NLP

TECHNOLOGY STACK

Some of the frameworks and tools that our development teams have used recently. A list that grows by the day.

RELATED RESEARCH

Relevant, interesting and current curated research content in the field.

Behavioral Biometrics’ Continuing Growth into Business Applications

October 27th, 2020|Categories: AI, NLP, Machine Learning, Mobile Development|Tags: |

With the current work climate accelerating towards remote offices and customer services, cybersecurity and, to be more specific, behavioral biometrics, is quickly becoming a focal point for any potential or recent innovation.

In fact, a recent report from Grand View Research concluded that the market size for behavioral biometrics is expected to grow at a compound annual growth rate of 24.5% from 2020 to 2027. The pandemic has accelerated interest in an already growing space, but should you start to consider adding behavioral biometrics to your service or business?

What is Behavioral Biometrics?

Behavioral biometrics is a form of cybersecurity that relies on the behavior of the user rather than static information or physical characteristics. Traditional security or static information for applications or consumer websites would include a password, phone number, or social security number. Physical characteristics would include security functions like fingerprint or face ID.

More recently, dual-step authentication is also a form of static information. For example, this would be when you’re asked to enter a password and also a numerical code sent to another device, like your phone, to confirm you’re signing in.

And while most of these types of security are relatively new, they are also still vulnerable. You’ve probably heard or read countless stories of data breaches even within companies keeping user information protected behind two-step authentication or face ID.

Behavioral biometrics is powered by artificial intelligence (AI) and works in the background of a user session and tries to identify if the person using the device or application is actually the normal user of that device. As an example, this version of biometrics can identify a user by the way the device is being held, the way a user normally types, or the way they navigate with the app’s interface. Behavioral biometrics can detect abnormalities between user interaction and automated/fraudster attacks.

How is Behavioral Biometrics Currently Being Applied?

In the past few years, the financial sector has seen the greatest investment in behavioral biometrics followed by insurance and eCommerce.

A top bank in the UK, for example, utilizes behavioral biometrics through industry leader Biocatch for its online banking app. Behavioral biometrics was able to alert the business in real-time of attempted fraud of almost 1.6 million pounds. The bank also utilized anti-malware and device recognition, but behavioral biometrics was the only security technology to alert the company of this attack.

Mastercard ran a report earlier in 2020 and revealed 7 out of 10 consumers believe the shift to digital payments is permanent. With this train of thought from the public, the growing need for investing in security measures to build trust with users is even more important now. Mastercard also began issuing contactless biometric bank cards back in 2019 and started trials for both debit and credit cards.

Why Should My Business Consider Frictionless Cybersecurity?

One of the more frustrating aspects of the common e-commerce website or mobile app is having to reset passwords or go through two-step authentication. These potential pitfalls of losing users to these frustrating and sometimes time-consuming tasks can ultimately lose business for good.

Behavioral biometrics removes any of those burdens and frictions on the user with a passive cybersecurity measure that also keeps the user safe. In addition, related traditional security costs tied to IT or help desks can also potentially be reduced when investing in behavioral biometrics.

More and more security companies are beginning to add behavioral biometrics solutions to their offerings, so now is the time to do research in this space. Just recently, private security company, Incognia, also added behavioral biometrics as an offering to their authentication services.

Is Behavioral Biometrics a Viable Option in 2020?

American consumers lost almost $17 billion last year because of identity fraud according to lead fraud and security analyst John Buzzard at Javelin Strategy & Research.

With the increased demand by consumers in wanting a contactless option for payments or e-commerce that also offers trusted security and frictionless user experience, behavioral biometrics is a potential, long-term solution for any company. And, with the growing competition in the market, there are more options than ever for a business of any size to do their due diligence in looking into cybersecurity that can ultimately save IT expenses and security costs.

Contact us at info@quantilus.com for a consultation and learn more about what Quantilus has to offer here.

Machine Learning: The Brain Behind Self-Driving Cars

October 26th, 2020|Categories: AI, NLP, Machine Learning, Emerging Tech - AR, Cloud, Blockchain|

Previously, we have explored machine learning and its impact on small businesses. However, machine learning is integrated in many aspects of life. Let’s switch gears and explore the applications of machine learning in the context of our lives as consumers. The automotive industry commonly uses artificial intelligence to optimize the design and manufacturing of vehicles released to the market. One area where machine learning plays a critical role is in the advent of self-driving cars. 

A handful of tech and car companies are currently racing to have their self-driving cars on the road first. The self-driving cars being developed rely on advanced technologies, including machine learning, to power these cutting-edge vehicles. So, before the day comes that you look in the rearview mirror and notice you are surrounded by self-driving cars, let’s go over the technology that propels these innovative cars.

 

Machine Learning to Drive Decisions

Machine learning is at the center of self-driving cars. For a self-driving car to drive on the road, it utilizes specially designed hardware and software. The hardware continuously collects data from the surrounding environment, while the software takes the collected data and sorts it. The software then relinquishes the data to machine learning algorithms to process the data and make decisions.

With machine learning algorithms, the more data the algorithm is exposed to, the more intelligent the algorithm becomes. This is because the algorithms take what they learn from prior experiences, draw conclusions based on the data, and then perform the action that will result in the best possible outcome.

 

What Data Does Machine Learning Look At?

Machine learning algorithms in self-driving cars are fed data and process it in real-time. The machine learning algorithms analyze data, including everything from recognizing upcoming stop signs to identifying deer on the side of the road and cars braking ahead. The vehicle then interprets this data and decides the best course of action. 

Also, it is important to note that machine learning is already utilized by certain cars on the market. For example, lane detection uses machine learning to monitor the lanes on the road. When machine learning detects the car is drifting out of its lane, it will either notify the driver or actually steer the car so it remains in the correct lane. 

 

Breaking It Down Further

We can further break down the use of machine learning in self-driving cars into smaller segments. These segments are object detection, object recognition, and movement prediction. When machine learning algorithms in self-driving cars analyze data, the first step is to recognize an object. Once an object is detected, the algorithms then need to determine what exactly the object is. Lastly, the algorithm needs to determine if the object will potentially get in the car’s path.

For example, if the car recognizes an object and then determines it is a gas station on the side of the road, the car knows it can continue driving since the store will not move into its path. However, if the vehicle recognizes an object and then determines it is a kid riding a scooter, the car decides that the kid may ride out in front of the vehicle, and the car will slow down.

 

Machine Learning for Prescriptive Alerts

Machine learning in cars is also used for the next level of predictive maintenance. Currently, most drivers experience notifications when their vehicle requires maintenance, such as the check engine light illuminating on the dashboard. However, machine learning takes it a step further by discovering budding issues before they happen. Additionally, machine learning can also provide prescriptive alerts. The purpose of prescriptive alerts is “to produce outcome-focused recommendations for operations and maintenance.” 

Essentially, prescriptive alerts determine when something is going to be an issue and then provides the solution, all before the issue actually arises. Examples of prescriptive alerts include notifying drivers when to change tires, replace brakes, and fix cooling systems. Prescriptive alerts and maintenance are extremely beneficial to the shipping industry and truck drivers. This is because drivers will not have to stop mid-trip to fix these issues which would delay the shipment, but instead can have their truck maintained ahead of time to avoid delays. Prescriptive alerts and maintenance are still new but are thought to be the next level of predictive maintenance.

 

The Dilemma with Self-Driving Cars

One of the most significant issues with self-driving cars is making sure they collect enough real-world experience and data on public roads. It is a catch-22 because on the one hand, self-driving cars need to be on public roads to collect data and be exposed to real-world situations to improve accuracy and safety. However, allowing self-driving cars on public roads also puts the public’s safety at risk by exposing them to the emerging technology.

 

Conclusion

Self-driving cars are not expected to be widely available and legal to the public for years to come. The leading companies currently building and testing these vehicles have predominantly already passed their initial predictions of when the cars will be ready. Therefore, they are still perfecting the technology behind the wheel (if there even is a wheel). However, without machine learning, self-driving cars probably would not have made it this far in their development. 

Machine learning is vital to the success and advancement of self-driving cars and is central to the vehicles’ operation. As advancements in machine learning are made, not only will self-driving cars improve, but so will countless other aspects of the publics’ everyday lives. From eCommerce to marketing and healthcare, machine learning will be innovating life as we know it. 

Interested in developing more about machine learning and how it can apply to your business? Contact us at info@quantilus.com for a consultation and learn more about what Quantilus has to offer here.

Natural Language Processing Explained: How Can it Impact My Business?

August 28th, 2020|Categories: AI, NLP, Machine Learning|Tags: , |

Computers were invented to interact with humans. These machines have always functioned by receiving commands from a user and performing a task accordingly. In the past, those tasks were relayed through code, or in the case of the earliest computers back in the 1970s, a punch card.

But times change, and as technology advances, so too does the means by which we communicate with it. That’s where Natural Language Processing comes into play.

Natural Language Processing, or NLP, allows computer programs to understand spoken or written language. It is an advancement in the field of artificial intelligence.

The main issue surrounding the inability of humans and computers to interact seamlessly was a language barrier. Machines do not speak the same language as us. They understand binary code, which is a series of millions of ones and zeroes that instruct a computer in completing their tasks. Systems have been set up where, with the press of a button or the click of a mouse, that computing language is relayed at the speed of light, allowing machines to understand our wishes.

That has all changed with the birth of NLP. We’ve all experienced advancements in this field. From Amazon’s Alexa, to Google Assistant, to Apple’s Siri, we now communicate with technology directly on an everyday basis.

“Alexa, order a new set of lightbulbs.”

“Ok Google, where is the best Chinese food near me?”

“Hey Siri, what song is this?”

Through the advancement of NLP, we’re always just a sentence away from the information and actions that we want.

But how does NLP work? What other uses does it have? What advantages can we expect to see from this technology in years to come?

How Does NLP Work?

NLP currently works through a process called deep learning.

Deep learning has the artificial intelligence look at data patterns to deepen its understanding of language. Huge amounts of labeled data are inputted to help the system identify relevant correlations.

Language is broken down into shorter elemental pieces in order to teach the machine to understand their relationships and how they work together. By doing this, the computer can ascertain the meaning behind a sentence.

Some of these data pieces include:

  • Categorization: This is a document summary based on linguistics. It includes indexing and searching, detection of duplicates, and content alerts.
  • Modeling and Topics: This helps machines understand the themes and meanings within a collection of text. They then take that meaning and approach it from an advanced analytical standpoint.
  • Context: Computers gain the ability to pull specific contextual information from the text.
  • Sentiment: Systems can understand the mood behind text. It analyzes the opinions expressed.
  • Speech-To-Text: This is a back and forth system that allows a machine to take a voice command and transform it into text. It also allows the computer to take written text and relay it vocally.
  • Summarization: Allows the computer to create a synopsis of a large text body.
  • Machine Translation: The computer can automatically translate text or speech from one language into another.

Deep learning represents a more fluid and intuitive approach to learning. By understanding the intention of the users, computers are able to learn language in the same way a child would. A human toddler listens to language, ascertains its meaning, and relays it back through a series of trial and error until fluency is achieved.

NLP works the same way, only in a much faster way.

What are the Uses of NLP?

Does NLP have a higher purpose beyond just telling you the movie times or reading off Yelp reviews?

Of course it does.

Search is the main function of NLP right now. We use many of the services mentioned above to find the information that we need. Whether that’s some arbitrary fact you and a friend are arguing over or a piece of information you need for a research project, searching via voice is far easier and more effective than manually inputting information.

When you ask your phone a question the machine is able to isolate the most important elements of your query. That’s why voice searches are usually so fast.

NLP can also be used to digitize hard copy information, making it easy to analyze and search for. Whereas once, a human operator would have to manually input all of the information, telling the computer what it all means and how to file it, NLP allows the system to understand the text on its own and file it accordingly.

Using the sentiment analysis that we discussed in the last section, businesses could have a better understanding of how their customers feel about their product or service. Computer systems can analyze large amounts of online comments and reviews to determine whether a business is succeeding or failing in the public eye.

In Conclusion

Natural Language Processing is the future of technology. We’ve already come so far since the creation of the first virtual assistant programs. Many of us entrust artificial intelligence with vital tasks like keeping track of our schedules. As we continue to deepen and enrich artificial intelligence programs, we can be sure that NLP will have a strong place in our daily lives for years to come.

Interested in implementing ML or AI in your business process? Contact us at info@quantilus.com for a consultation and learn more about what Quantilus has to offer here.

3 Ways to Implement Machine Learning and AI into Small Businesses

August 26th, 2020|Categories: AI, NLP, Machine Learning, Mobile Development, Web Development, Marketing, SEO|Tags: , , , , , |

Earlier this month, John Giannandrea, Apple’s head of artificial intelligence (AI), gave insight into how Apple is leveraging machine learning (ML) within their iOS and the future of machine learning at Apple. Anything from language translation to only sorting photos on your phone into pre-made galleries is made possible with machine learning. These various types of machine learning applications are becoming more and more common and essential.

However, a common misperception of AI and machine learning is that these advanced and sophisticated technologies are only for big brands with budgets that allow for experimentation and implementation.

A small business owner reading a report like Gartner’s 2019 survey of CIO’s would find that although 37% of organizations have installed some form of AI or machine learning, most of the CIO’s interviewed were from large brands. This type of survey may further intimidate small and medium-sized business owners into thinking the growing age of AI and machine learning isn’t ready for their company yet. But, you’d be surprised by how easily small companies can adopt cutting-edge technology without having to rely on an extensive budget.

Marketing

For small businesses, ML software as a service can be a great tool to utilize, especially in consumer and B2B marketing spaces. In fact, 40% of marketers prioritize AI and machine learning more than any other department and consider them critical to their success. 

One familiar and accessible ML tool marketers at small businesses can leverage is a chatbot tool for their website. Chatbots now utilize natural-language processing (NLP) that can interact conversationally with website visitors and collect information like preferences visitors have as they browse a website.

Chatbot tools like Botsify also allow for integration with several services and offer an easy interface to help customize your company’s brand into their templates.

In terms of ML digital marketing tools, a chatbot is just one of the many corners that can be explored. Also, consider implementing machine learning marketing applications for email marketing, ad targeting, voice search, or predictive analysis to help create your campaigns reach new, multiple touchpoints.

Security

Like marketing, cybersecurity is witnessing a fast-growing trend with investments into machine learning tools to help protect their own company and customers. ABI Research estimates ML, AI, and big data spending will increase to $96 billion by 2021.

Machine learning technology can track users’ patterns and make assessments of these patterns, such as the iPhone creating galleries from related pictures, as previously mentioned. This technology can be applied for security responsibilities by implementing security algorithms into something like your mobile application.

If your business relies on the consumer making financial transactions, it’s the company’s responsibility to keep any entered information secure.

Biocatch, for example, utilizes behavioral biometrics, a machine learning technology that tracks user behavior within an application as another form of security. Behavioral biometrics can identify when a different person uses an application by the way they move around the app or the way they type.  

Sometimes even just adding a single line of code can improve your mobile application’s security and add another form of authentication working in the background.

Accounting

Another aspect of any business that requires managing a ton of data, repetitiveness, and predictably is the accounting department, which means machine learning for accounting tasks is a relationship that makes too much sense to not explore.

The future of accounting is heavily entangled with AI and ML, and back in 2018, a report suggested tasks like taxes and payroll would eventually become fully automated. In 2020 machine learning is now applied to generating and processing invoices, even including specific requirements with each task.

Xero, a New Zealand-based accounting software company, provides a cloud accounting software that creates and processes invoices for small businesses. The machine learning aspect in the software enables the creation of invoices based on past behaviors for customers.

Whether you leverage machine learning through marketing, security, or accounting, the decision ultimately depends on where you possibly see the value and ROI of any machine learning tool. For a small-to-medium-sized business, where that value is found can vary.

More and more companies are tapping into or at least exploring how machine learning can impact their business as the machine learning market is expected to grow at 42% CAGR by 2024. While adoption of AI and ML may be daunting, machine learning models are proving to show increasing value to any business size.

Interested in implementing ML or AI in your business process? Contact us at info@quantilus.com for a consultation and learn more about what Quantilus has to offer here.

Bitnami