Author: altilia

The explosion of interest and publicity in Artificial Intelligence in recent months has come from the advent of Large Language Models, specifically OpenAI’s ChatGPT, which set the record for the fastest-growing user base in January.

Suddenly it seems like everyone is fascinated by the coming surge of AI with new applications, creating excitement and fear for the future.

When Google’s so-called “Godfather of AI” Dr Geoffrey Hinton warned about “quite scary” dangers, it made headlines around the world.

Behind the hype

So, it is important to understand what is behind the hype and see how it works and what your organization can use to build future value.

This blog is split into two: first we learn about Natural Language Processing, the branch of computer science concerned with giving machines the ability to understand text and spoken words in much the same way humans can.

And then we will go deeper on Large Language Models (LLMs), which is what ChatGPT and others like Google’s Bard are using.

NLP combines computational linguistics with statistical, machine learning, and deep learning models to enable computers to process human language in the form of text or voice data and to ‘understand’ its full meaning, complete with the speaker or writer’s intent and sentiment.

NLP drives computer programs that translate text from one language to another, respond to spoken commands, and summarize large volumes of text rapidly—even in real time.

There’s a good chance you’ve interacted with NLP in the form of voice-operated GPS systems, digital assistants, speech-to-text dictation software, customer service chatbots, and other consumer conveniences.

But NLP also plays a growing role in enterprise solutions that help streamline business operations, increase employee productivity, and simplify mission-critical business processes.

There are two sub-fields of NLP:

• Natural Language Understanding (NLU) uses syntactic and semantic analysis of text and speech to determine the meaning of a sentence, similarly to how humans do it naturally. Altilia uses Large Language Models for this.
• Natural Language Generation (NLG) enables computers to write a human language text response based on data input. ChatGPT uses LLMs for NLG.

Large Language Models (LLMs)

LLMs are a relatively new approach where massive amounts of text are fed into the AI algorithm using unsupervised learning to create a “foundation” model, which can use transfer learning to continually learn new tasks.

The key is using huge volumes of data. The training data for ChatGPT comes from a diverse set of text sources, including billions of web pages from the internet, a huge number of books from different genres, articles from news websites, magazines and academic journals and social media platforms such as Twitter, Reddit and Facebook to learn about informal language and the nuances of social interactions.

The model is then able to predict the next word in a sentence and generate coherent text in a wide range of language tasks.

Altilia does exactly the same, but uses this capability to provide enterprise tools for specific business use cases.

Technology breakthrough

Overall, NLP is the core technology to understand the content of documents. LLMs are a breakthrough in the field as they allow a shift from where an NLP model had to be trained in silos for a specific task to one where LLMs can leverage accumulated knowledge with transfer learning.

In practice, this means we can apply a pre-trained LLM and fine-tune it with a relatively small dataset to allow the model to learn new customer-specific or use-case specific tasks.

We are then able to scale up more effectively, it can be applied more easily for different use cases, leading to a higher ROI.

For more information on how Altilia Intelligent Automation can support your organization to see radical improvements in accuracy and efficiency, schedule a demo here.

The rise of Artificial Intelligence has been the talk of the business world since the emergence of ChatGPT earlier this year.

Now executives around the world find themselves in need of understanding the importance and power of Large Language Models in delivering potentially ground-breaking use cases that can bring greater efficiency and accuracy to mundane tasks.

Natural Language Generation (NLG) enables computers to write a human language text response based on human generated prompts.

What few understand is that there is still a deep flaw in the ChatGPT technology: up to 20-30% of all results have inaccuracies, according to Gartner.

What Gartner have found is that ChatGPT is “susceptible to hallucinations and sometimes provides incorrect answers to prompts. It also reflects the deficiencies of its training corpus, which can lead to biased or inappropriate responses as well as algorithmic bias.”

To better understand this, it’s key to consider how LLMs work: hundreds of billions of pieces of training data are fed into the model, enabling it to learn patterns, associations, and linguistic structures. This massive amount of data allows the model to capture a wide range of language patterns and generate responses based on its learned knowledge.

However, as vast training data can be, the model can only generate responses as reliable as the information it has been exposed to. If it encounters a question or topic that falls outside the training data or knowledge cutoff, responses may be incomplete or inaccurate.

For this reason, and to better understand how best to use LLMs in enterprise environments, Gartner outlined a set of AI Design Patterns and ranked them by difficulty of each implementation.

We are delighted to share that Altilia Intelligent Automation already implements in its platform two of the most complex design patterns:

• LLM with Document Retrieval or Search
  • This provides the potential to link LLMs with internal document databases, unlocking key insights from internal data with LLM capabilities
  • This provides much more accurate and relevant information, reducing the potential for inaccuracies due to the ability to the use of retrieval.
• Fine-tuning LLM
  • The LLM foundation model is fine-tuned using transfer learning with an enterprise’s own documents or particular training dataset, which updates the underlying LLM parameters.
  • LLMs can then be customized to specific use cases, providing bespoke results and improved accuracy.

So, while the business and technology world has been getting excited by the emergence of ChatGPT and LLMs, Altilia has already been providing tools to enterprises to leverage these generative AI models to their full potential.

And by doing so, thanks to its model’s fine-tuning capabilities, we are able to overcome the main limitation of a system like OpenAI’s ChatGPT, which is the lack of accuracy of its answers.

For more information on how Altilia Intelligent Automation can help your organization, schedule a free demo here.

Welcome to our third blog of a series uncovering the key components of Artificial Intelligence to provide greater understanding for business leaders who may currently have FOMO (Fear Of Missing Out) from the blizzard of acronyms and hype.

Here, we look at Computer Vision, one of the main applications of AI where computers can be made to gain high-level of understanding from digital images or videos.

Critically, Computer Vision is concerned with automatic extraction of data, enabling documents that have handwriting and random layouts to become machine-readable.

Huge data volumes

Computer Vision needs a lot of data to be able to distinguish and recognize images.

In a way, it looks like a jigsaw puzzle where you assemble all the scattered tiles to make an image. Neural networks for CV work on the same principle.

Yet the computer does not have the final image, but it is fed hundreds, if not thousands of related images that train it to recognize specific objects.

To identify a cat, the computer would not be shown individual elements such as ears, whiskers, tail etc, but millions of pictures of cats so that it can model the features of our feline friends.

CV is used for visual surveillance, medical image processing for patient diagnosis and navigation by autonomous vehicles.

But in Altilia’s development of Intelligent Document Processing (IDP), CV has several key roles to play.

With Optical Character Recognition (OCR) and Intelligent Character Recognition (ICR), we are able to convert scanned documents into machine-readable PDFs and with Handwritten Text Recognition (HTR) are incorporate items such as signatures.

End goal

The end goal of an IDP solution is to extract meaningful information that are “hidden” in unstructured texts and documents, so we need to first break words down in a way that a machine can understand.

This is especially relevant when the documents that need to be processed are (low quality) scans such as contracts, forms, invoices or ID cards.

We then need to apply OCR to recognize both printed and handwritten text, using smaller units called tokens. To each token is added metadata, which is useful later in a search engine.

In IDP, it is useful to distinguish a photo from text and to tag elements such as signatures, stamps and markings, saving human labor time by automating checks such as whether a contract is signed and marked.

Finally, we focus on document layout analysis so that unsorted documents can be classified and then we can apply different machine learning algorithms and branch out different ML pipelines.

These core capabilities allow Altilia’s solution to work as a general purpose platform, rather than a point solution for specific document types and formats. We have also developed a patented solution for document layout analysis.

For more information on how Altilia Intelligent Automation can help your organization, schedule a free demo here.

In our second blog of this series, where we unlock the lexicon of Artificial Intelligence for business leaders currently being overwhelmed by the hype of ChatGPT, we will focus on Machine Learning (ML).

What is Machine Learning?

People throw the terms machine learning and AI together and interchangeably, but they don’t mean the same thing. ML is a subset of AI that uses computers to learn or improve performance based on the data they use.

It’s a fascinating concept, straight out of science fiction: a computer uses algorithms to learn from the data provided. The more it develops, the more it learns: the more data it is fed, the better it gets.

It is where the concerns come that computers can become “more intelligent” than their human masters.

The reason ML has become more successful and prominent in the past decade, is the growth in volume, variety and quality of both public and privately-owned data, the availability of cheaper and more powerful data processing and storage capabilities.

Essentially ML models look for patterns in data and draw conclusions, which is then applied to new sets of data. They are not explicitly directed by people, as the machine learning capabilities develop from the data provided, particularly with large data sets. The more data used, the better the results will be.

So, where AI is the umbrella concept of enabling a machine to sense, reason or act like a human, ML is an AI application that allows computers to extract knowledge from data and learn from it autonomously.

How to train ML models

The key to machine learning (as much else in life) is training. ML computers need to be trained with new data and algorithms to obtain results.

Three training models are used in machine learning:

• Supervised learning maps in a specific input to an output using labelled/structured training data. Simply, to train the algorithm to recognize pictures of cats, it feeds it labelled pictures of cats.
• Unsupervised learning is based on unstructured (unlabelled) data, so that the end result is not known in advance. This is good for pattern matching and descriptive modelling. For example, Altilia uses Large Language Models (LLMs) as its foundation, which are trained on huge datasets using unsupervised learning.
• Reinforcement learning can be described as “learn by doing”. An “agent” learns to perform a task by feedback loop trial and error until it performs within the desired range, receiving positive and negative reinforcement depending on its success. Altilia often uses Human-in-the-Loop (HITL) reinforced learning in its Altilia Review module.
• Transfer learning enables data scientists to benefit from knowledge gained from a previous model for a similar task, in the same way that humans can transfer their knowledge on one topic to a similar one. It can shorten ML training time and rely on fewer data points. Altilia uses this technique to fine-tune pre-trained Large Language Models (LLMs) on a dataset provided by the client. We will focus on LLMs in a future blog.

Why not schedule a demo with Altilia to learn more about how we can help transform your organization? Click here to register. 

The emergence of ChatGPT in recent months has prompted an unprecedented explosion of public hype and interest in Artificial Intelligence and its potential future uses.

Following industry luminaries such as Bill Gates describing AI as the most important technological advance in decades, there has been a torrent of predictions on how it will change the world of work.

Here at Altilia we have been building an AI-based platform, Altilia Intelligent Automation, that is at the forefront of a new revolution based around Intelligent Document Processing (IDP).

It is clear, however, that industry leaders anxious to understand this new phenomenon, are overwhelmed by the language, acronyms and blizzard of nascent technologies which leave them baffled and unsure on how to get started.

Our aim over a series of blogs is to unpack the lexicon of AI and explain what it all means, a useful, insightful and simplified guide for organizations interested in how AI will affect them.

What is Artificial Intelligence?

The term artificial intelligence (AI) is used loosely to refer to applications that perform complex tasks that previously required human intervention, such as communicating with customers online or playing chess.

The term is often used interchangeably with the terms machine learning (ML) and deep learning, although they (as we will discover) have different meanings.

Many companies are investing significantly in data science teams to take full advantage of AI, combining statistics, computer science and business knowledge to extract value from various data sources and enable problem-solving. Algorithms seek to create expert systems which make predictions or classifications based on input data.

Advanced functions can include the ability to see, understand and translate spoken and written language, analyze data, make recommendations and classify both structured and unstructured data.

Computers and machines are developed with the ability to reason, learn and act in a way that would normally require human intelligence, often at a scale that exceeds or speeds up what humans can deal with.

Benefits of AI:

• Automation. AI can automate workflows and processes or work independently and autonomously from a human team.
• Reduce human error. AI can eliminate manual errors in data processing, analytics, assembly in manufacturing, and other tasks through automation and algorithms that follow the same processes every single time.
• Eliminate repetitive tasks. AI can be used to perform repetitive tasks, freeing human capital to work on higher impact problems.
• Fast and accurate. AI can process more information more quickly than a human, finding patterns and discovering relationships in data that a human may miss.
• Accelerated research and development. The ability to analyze vast amounts of data quickly can lead to accelerated breakthroughs in research and development.

Altilia are world experts in Intelligent Document Processing (IDP), so let’s have a look at how it builds on AI capabilities and developments to enhance organizational efficiency and accuracy.

What is Intelligent Document Processing?

At its most simplistic level, Intelligent Document Processing (IDP) converts unstructured and semi-structured data into structured usable information, thus enabling layers of automation to document-centric business processes.

As an example, many mortgage forms may be filled in with (unstructured) hand-written answers by an applicant, which would need human intervention to input that information into a financial services company’s systems.

IDP uses AI (and other technologies) to extract that information in a usable form, thus reducing time and manual labor.

Which AI fields are relevant for IDP?

A human reading and understanding a document needs to:

• Have visual perception to recognize images, symbols and writing
• Have a comprehensive understanding of the document’s language
• Be able to understand new information and learn concepts
• Be able to memorize concepts and the relations between them.

AI, therefore, needs to emulate these abilities to be capable of reading like a human – and for IDP, these are the most relevant application fields:

• Computer Vision is the AI field that emulates human visual perception
• Natural Language Processing (NLP) focuses on developing algorithms for general understanding of a language
• Machine Learning (ML) is focused on training a machine to learn and conceptualise information as models
• Knowledge Representation is focused on representing information in a way that is functional, not just to memorize the content, but relations between them.

We will continue this series explaining the lexicon of AI in the coming weeks. For more information on how Altilia can support your business, schedule a demo here .

Intelligent Automation is being embraced by many organizations to boost revenues, increase efficiency, and improve customer experiences.

As an evolution of Robotic Process Automation (RPA), Intelligent Automation utilizes a Composite AI approach to leverage multiple techniques, such as machine learning (ML), computer vision (CV), and natural language processing (NLP) to handle complex processes and support business decision making.

Intelligent Automation has the potential to revolutionize the workplace, not only by freeing up time human workers dedicate to repetitive tasks but also helping companies to set new strategies and to develop new innovative products and services.

Unfortunately, this potential is still largely untapped, as most companies have yet to fully embrace it. Many opportunities are still left on the table.

Some of the best candidates for automation are business processes dealing with structured data (both digital and non-digital) with clearly defined stages; processes with seasonal spikes that cannot be easily met by a manual workforce (such as policy renewals, premium adjustments, claims payments); processes with strict Service Level Agreements (SLAs), requiring fast turnaround times (like transaction registration, order fulfilment, etc.).

Some notable use case applications also include:

• Contract management: Intelligent automation can help to simplify all the steps needed to draft, send, redline, and execute contracts.
• Sales processes: Here automation can provide first level assistance to prospects by providing direct answers to FAQs.
• Quality and security compliance procedures: Intelligent automation can be leveraged to collect and pre-process data to perform initial assessments before involving a human analyst.
• Finance: It can be used to streamline the procure-to-pay (P2P) process.
• HR: Many tasks in the Hire-To-Retire (H2R) process can be automated.
• Customer service: Many contact center tasks can be automated, as virtual agents can handle routine tasks, freeing up human agents to handle more complex customer inquiries, with great benefits for the customer experience.

Intelligent Automation has the potential to change the architecture of work by breaking it into smaller tasks and making it more events driven.

In the future, fewer tasks will need to be handled manually, leaving workers with more time to focus on digital learning and knowledge work, including learning how to proactively develop solutions using low-code platforms and tools.

The use of intelligent automation can boost the efficiency and effectiveness of business processes, enabling companies to meet higher standards of both customer and employee satisfaction.

However, as organizations increasingly adopt intelligent automation technologies, they need to be aware of the potential risks and hazards to avoid.

One of the main pitfalls is ignoring the importance of change management.

This could result in long-term issues if organizations overlook the importance of having their people aligned with their overall goals.

Additionally, employees may push back when automating processes that were previously carried out by a team of people. It is important to understand and take into account the concern and perceived lack of control when processes are suddenly automated.

Lastly, in order for any automation project to be successful it necessary to set clear goals in advance and to establish key metrics to evaluate the impact of the initiative and to measure the ROI.

Biggest challenges

Scaling intelligent automation is one of the biggest challenges for organizations: therefore, it is crucial for companies to be clear about the strategic intent behind their initiatives.

Success often depends on the ability of the organization to put “people’s needs” first: this means introducing new technologies in a way that is helpful and involves minimal disruption, and addressing real issues related to skills, roles and job content.

In other worlds, companies should not approach the problem like a race to be the first to introduce the latest technology and neither as a way to substitute people in managing operations.

Intelligent Automation can rather realize its full potential when it implemented as a means to support and augment human employees.

This allows a correct balance to be found between automation and manual control, leading to the best results when determining the success of a project.

The right approach

If organizations take the right approach, employees can also positively embrace automation and the opportunities it creates, rather than resist it.

In fact, automation can free up time on repetitive work, helping employees to focus on high-priority tasks and allowing them to move from administrative management to higher-value contributions.

Forward thinking organizations will try to shift workers into new roles as the current ones get progressively automated; this way employees can learn about new technologies and how they can be leverage them.

With this vision, upskilling programs, both internal and external, will be more and more a differentiating factor for companies to attract and retain the best talents.

For more information on how Altilia can help you gain the full benefits of Intelligent Automation, contact us here.

As we explored in our previous blog, demand for AI-based solutions is growing year on year as organizations adopt and experiment with artificial intelligence to drive competitiveness and efficiency in a tough business environment.

We examined the key talent and skills gap challenge that companies, particularly small and medium enterprises, are facing as the explosion of AI projects outstrips the supply of qualified data experts required.

Now we want to go deep on another significant issue that has been identified by business leaders and consultants as a blocker to growth. And also show how Altilia can offset this challenge.

Managing Machine Learning (ML) Operations

When companies consider introducing new business process automation solutions, they often underestimate the complexity of managing the lifecycle of an AI project.

ML Ops are defined as a set of practices that combine the implementation of Machine Learning, DevOps, and Data Engineering models, with the goal of developing, releasing, monitoring, and the scaling into production high quality ML systems. This is done by leveraging organizational, cultural, and technological aspects that support the governance and automation of the ML model lifecycle.

Properly managing ML Ops is a major challenge even for large and structured companies.

In fact, according to a recent survey by the Artificial Intelligence Observatory of the Polytechnic University of Milan, out of a sample of 80 large companies operating in Italy that have already started AI projects, 71% say they have not yet introduced ML Ops as a structured practice.

Among the stated reasons for this delay stands out the lack of time they devote to AI initiatives compared to traditional business activities (43%) and the lack of internal expertise to manage ML Ops (33%), an issue which brings us back to the challenge discussed in our last blog.

Deloitte identified this ML Ops issue in their State of AI report: “Despite evidence that establishing clear processes and redefining roles to deliver quality AI results in improved outcomes, there has been little growth in the market in terms of adopting such practices, according to survey respondents.

In both the fourth and fifth editions, just one-third of respondents reported that their companies are always following MP Ops, redesigning workflows and documenting AI model life cycles.”

The ML Ops cycle in its complexity could be broken down in three distinctive parts:

• The Business process analysis/understanding part refers to the mapping of internal processes, systems, and any other key element to understand the “problem” to be solved: such as defining requirements, objectives, key outcomes, and formulating use case hypotheses for ML applications.
• The Data & Model Preparation part concerns the collection and refinement of data to train the ML model; the selection of the ML model with the best fit to solve the problem; the testing of different algorithms, features and hyperparameters; keeping track of all experiments, and maximizing code reusability.
• Continuous Operations & Improvement refers to the continuous update of the solution with each product release; the continuous training of the algorithm, with the storage and processing of new data to update the model; the continuous monitoring of the model to keep track of changes in performance and accuracy.

How can Altilia help to overcome the ML Ops challenge?

Altilia’s approach is offering a comprehensive platform to manage the whole AI implementation lifecycle with an end-to-end approach.

The platform is built overtime and improved upon experience by Altilia, to hide the complexity of managing ML Ops, and to simplify the training and the optimization of AI models over time.

This eliminates the need to dedicate an internal team (with limited experience) just to maintain an AI implementation project, to monitor its performance over time and to manage its infrastructure.

Additionally, the automation challenges faced by businesses are rarely completely unique. Using a platform approach, Altilia can replicate pre-tested use case specific solutions and adapt them to perfectly fit the needs of the customer.

This greatly simplifies the process analysis, reducing the uncertainty when defining implementation goals, a leading to a greater confidence about the expected outcomes.

Altilia simplifies data and model preparation by giving access to library of generalized pre-built AI models, including Large Language Models (LLMs) that can be trained and adapted, according to the customer’s specific requirements.

The training of models is simplified, thanks to our document annotation interface, that allows users to easily generate examples for the system to process.

Lastly, Altilia facilitates the monitoring and fine-tuning of AI models to increase their accuracy over time and prevent data drift.

Contact Altilia here to learn how our unique AI technology platform can help your organization overcome these challenges.

The demand for AI-based solutions in the enterprise is ever increasing as more and more companies view artificial intelligence as a solution to automate core business process, driving operational excellence and accelerating digital transformation.

Top-level research from the Artificial Intelligence Observatory of the Polytechnic University of Milan shows that, from a sample of 200 large organizations operating in Italy, 59% have already started executive projects in the AI field, 77% of which are running more than one project in parallel.

According to a recent survey by PwC in US, on a sample of 1000 large organizations, 72% have already integrated or are looking to integrate AI technologies into their operations.

Full-scale deployment

Deloitte’s significant State of AI study found that 79% of business leaders surveyed reported full-scale deployment for three or more types of AI applications – up from 62% in the previous year. And 76% of those leaders expect AI investments to increase in the next year.

But not everything in the AI garden is rosy, particularly for small and medium sized organisations.

There are many challenges affecting the effective use of artificial intelligence which can hinder the ability to tap into its potential and lead to a growing gap between those companies that have the capability and those who do not.

Over two blogs, we will focus on the two big challenges facing organisations planning to use AI – and how Altilia can make the difference.

1. The AI skills shortage
2. The complexity of managing Machine Learning Operations

The AI skills shortage

McKinsey’s State of AI 2022 Survey called out the difficulties of hiring AI-related roles, with 78% of respondents citing difficulty with hiring AI data scientists, 72% struggling to hire data architects, 70% ML engineers and 69% data engineers.

Perhaps surprisingly, they state that finding AI hires is perceived as more difficult than three years ago, with 57% saying that the hiring of data engineers is more difficult, 52% for data architects and 48% of ML engineers.

Clearly, demand for top AI talent is outstripping supply.

Among the limiting factors, the skills problem emerges as one of the most relevant. In addition to programming and software development skills, the planning and life-cycle management of an AI project, requires highly specialized skills that are still scarcely available in the labour market.

These technical skills include:

• Skills needed to identify and prepare datasets that will be used by AI algorithms.
• Mathematical and statistical skills that enable the solving of complex problems related to AI applications.
• Skills needed to implement Machine Learning and Deep learning models (e.g., Supervised & Unsupervised Learning techniques).
• Skills related to data storage infrastructure, cloud services, and integration of AI systems with other enterprise software and applications.

In addition, since AI automation projects are ultimately aimed at boosting operational business processes, it is crucial to create a match between AI skills and all the skills involving the application of specialized domain-specific knowledge, methodologies, and operational techniques for the operation of specific areas, such as Supply Chain, Marketing and Sales.

For this reason, a core challenge is to bridge the gap between technical AI skills and domain-specific competences within companies and bringing core AI skills to figures not traditionally related to engineering and software development.

How can Altilia help to overcome the skills gap challenge?

Altilia prioritizes AI democratization as the core of its approach: for this reason, the Atlilia Intelligent Automation platform is designed to maximize intelligent process automation capabilities while minimizing the complexity of managing AI algorithms for the end user.

• Thanks to the use of pre-built algorithms, based on transformer models, users can leverage AI capabilities even without the in-depth mathematical and statistical knowledge related to AI applications.
• Thanks to the use of a no-code visual interface for document annotation and review, users can autonomously prepare the dataset that will be used by AI algorithms, even without the technical skills required to implement machine learning or deep learning models.
• Altilia’s IPaaS cloud infrastructure, equipped with pre-built software connectors, greatly simplifies software interoperability, and removes the complexity of managing computing resources and data storage.

In addition, the Altilia Intelligent Automation platform is designed for to reduce the learning curve for business-domain experts to understand and apply AI algorithms for business automation.

In fact, the Altilia Labels module for document annotation, allows business-line users to intuitively transfer their knowledge into algorithms, to perfectly fine-tune the solution according to real business and process needs.

The Altilia Reviews module provides a tool to review and validate the results produced by the algorithms, this is a crucial feature to guarantee AI explainability and to allow business-line users to effectively govern the process.

Contact us here to discuss how Altilia can support your organization’s drive for business optimization through AI.

And read our next blog on the challenge of managing ML Ops.

As 2022 is coming to an end, we want to take a step back and have a look at the state of the industry of AI for business automation.

McKinsey has just published their “State of Artificial Intelligence 2022” report, based on a survey of from 1,492 participants representing the full range of regions, industries, company sizes, functional specialties, and tenures.

The report demonstrates the growth and importance of AI to organizations, with a set of companies seeing the highest financial returns from AI continuing to pull ahead of competitors.

The results show these leaders making larger investments in AI, engaging in increasingly advanced practices known to enable scale and faster AI development, and showing signs of faring better in the tight market for AI talent.

There were two key takeaways from this important report:

1. AI adoption is increasing, and the market is growing.
2. Emerging AI leaders are reaping the rewards.

AI adoption is increasing, and the market is growing

Global AI adoption is increasing: companies adopting AI in at least one business area increased by 2.5 times since 2017. The average number of AI capabilities used by organizations has increased from 1.9 in 2018 to 3.8 in 2022.

Investments in AI have also increased: % of companies whose digital budget went for more than 5% to AI increased from 40% in 2018 to 52% in 2022.

Most used embedded capabilities:

• Robotic Process Automation 39%
• Computer Vision 34%
• Natural Language Processing 33%
• Deep Learning 30%
• Knowledge Graphs 25%
• Reinforcement Learning 20%
• Transformers 11%

Altilia is exceptionally well positioned, since its offer is strong for both widely adopted embedded capabilities (RPA, Computer Vision, NLP) as well as new emerging capabilities whose value is not yet fully understood by the market but will become critical in the coming years (Knowledge Graphs, Reinforced Learning, Transformers).

With companies starting to understand the cross-functional potential of AI automation application, this also puts Altilia in a good position as the Altilia Intelligent Automation platform is designed to let customers leverage the same infrastructure to implement multiple use cases. This helps to increase the ROI of the platform over time.

Emerging AI leaders are reaping the rewards

McKinsey recognizes a group of “AI high performers”, a group of companies that are being particularly effective with the implementation of AI projects. Companies whose EBIT have increased more than 20% from the use of AI (8% of total respondents).

These high performers are more likely than others to follow ‘frontier’ core practices that unlock value:

• Adopting practices for AI development at scale (AI Ops).
• Embracing modular AI architecture to rapidly accommodate new AI applications.
• Introducing quality control procedures to provide high-quality data to feed AI algorithms.
• Leverage emerging low-code and no-code programs to speed up the creation of AI applications.

These AI high performers are also more likely than other companies to report that they engage in practices to mitigate AI related risks, testing the validity of models and monitoring them over time for potential issues.

Altilia is again well positioned since Altilia Intelligent Automation is designed to let customers ‘outsource’ the complexity of AI Ops, a topic we will investigate further in a subsequent blog post.

Our platform has a modular architecture and it allows to simplify the deployment of multiple AI applications, and it’s also designed as a low-code/no-code platform, to simplify the creation of AI applications.

It is important to understand that Altilia Intelligent Automation combines the use of Large Language Models (LLMs) with no-code technology, and this constitutes a breakthrough.

LLM are already popular and publicly discussed, but those are general purpose technology that doesn’t perform very well for specific business applications.

By utilizing Altilia’s no-code approach to AI models training, these ‘general purpose’ technologies can be fine-tuned by enterprises, to turn them into a ‘domain-specific’ technology.

This will allow businesses of all sizes to experience the benefits of LLM technology in their specific business applications, allowing them to obtain returns on AI investments, comparable to the “high performers” mentioned by McKinsey.

The magical promise of artificial intelligence is that it learns as it goes along and develops additional data insights that enable your organization to make rapid progress.

Now Adaptive AI has been identified by Gartner as one of the top 10 emerging trends for 2023, taking AI capabilities to the next level where it is able to absorb learnings even as it’s being built.

They estimate that by 2026, enterprises that have adopted AI engineering practices to build and manage adaptive AI systems will outperform their peers in the operationalizing AI models by at least 25%.

In a recent Gartner article, they write: “Adaptive AI brings together a set of methods (i.e. agent-based design) and AI techniques (i.e. reinforcement learning) to enable systems to adjust their learning practices and behaviors so they can adapt to changing real-world circumstances while in production.”

What is Adaptive AI?

So, what is Adaptive AI and how does it differ to our current understanding of how AI works?

The key is that Adaptive AI can revise its own code to adjust for changes that weren’t known or predicted when the code was first written, enabling adaptability and resilience to be built into the design so that it can react immediately to changes.

It means that the “learning” phase of a traditional AI system can be bypassed so that the AI is effectively learning whatever is happening.

The value of operationalized AI lies in this ability to rapidly develop, deploy, adapt and maintain AI across different environments in the enterprise.

AI models with this self-adaption built in can develop quicker and with less errors. It creates a faster and more superior user experience by adapting to changing real-world situations.

Altilia at the forefront

Altilia is at the forefront of this approach to optimizing artificial intelligence to take it to the next level.

In our platform, we use reinforcement learning to improve the accuracy of our machine learning models over time.

Also, as a method, we have designed a human-in-the-loop feedback cycle that allows users to trace back the extracted data points to the original source (i.e. the exact position within the document).

This allows them to validate data, and the resulting feedback is taken into consideration to re-train the AI model. In this way the model can improve their accuracy over time and prevent data drifting.

It also means that if the format or layout of the processed documents is gradually modified over time, the algorithm is capable of adapting without the need to refactor our solution.

Why not schedule a demo of Altilia’s ground-breaking AI intelligent document processing solutions and see how it can revolutionize your organisation’s use of your data?

You can sign up here for your demo.