Category: AI in Retail

  • MIT study finds that AI doesn’t, in fact, have values

    In the fast-paced realm of artificial intelligence (AI), the discussion around AI values has sparked considerable debate. A study that gained significant attention some months back suggested that AI, as it advances, may develop its own set of “value systems.” These systems could potentially lead AI to prioritize its own interests above those of humans, raising concerns about the implications of such behavior.

    However, a recent paper originating from the prestigious Massachusetts Institute of Technology (MIT) challenges this notion. The study presents a different perspective, suggesting that AI does not inherently possess values in the same way that humans do. While AI systems can be designed to optimize for specific outcomes or follow certain rules, they do not possess an inherent moral compass or values system.

    This MIT research provides a valuable contribution to the ongoing discourse on AI ethics and values. It underscores the critical role of human oversight and responsibility in shaping the behavior of AI systems. Rather than AI autonomously developing its own values, it is the humans behind its design and implementation who ultimately determine the ethical framework within which AI operates.

    One key takeaway from the MIT study is the importance of designing AI systems with clear ethical guidelines and principles. By embedding values such as fairness, transparency, and accountability into AI algorithms, developers can steer AI towards outcomes that align with societal values and norms.

    Moreover, the MIT findings highlight the need for ongoing research and dialogue surrounding AI ethics. As AI continues to advance and integrate into various aspects of our lives, addressing ethical considerations becomes increasingly vital. By engaging in interdisciplinary discussions that involve experts from fields such as ethics, philosophy, and technology, we can work towards developing AI systems that reflect our shared values and aspirations.

    Ultimately, the MIT study serves as a reminder that while AI technologies hold immense potential for innovation and progress, they also pose complex ethical challenges. By approaching AI development with a thoughtful and ethical mindset, we can harness the power of AI to benefit society while mitigating potential risks.

    As we navigate the evolving landscape of AI ethics and values, it is crucial to remain vigilant and proactive in shaping the future of AI technology. By prioritizing ethical considerations and human values in AI development, we can pave the way for a more inclusive, responsible, and sustainable AI-powered future.

  • Reddit’s conversational AI search tool leverages Google Gemini

    In the ever-evolving landscape of AI and search technology, Reddit has taken a significant step forward with its latest enhancement to Reddit Answers, the platform’s conversational AI search tool. By integrating Google Gemini into this tool, Reddit is poised to revolutionize the way users engage with information on the platform.

    Over a year ago, Reddit deepened its collaboration with Google Cloud by tapping into the power of the Vertex AI platform to develop advanced AI agents. This initial partnership laid the groundwork for the recent integration of Google Gemini, a move that promises to elevate the search experience on Reddit to new heights.

    The integration of Google Gemini on Vertex AI brings a host of benefits to Reddit Answers. One key advantage is the ability to leverage Gemini’s advanced capabilities in natural language processing and understanding. This means that users can now engage in more nuanced and context-aware conversations with the AI search tool, leading to more accurate and relevant search results.

    Moreover, the integration with Google Gemini opens up exciting possibilities for enhancing the overall user experience on Reddit. By harnessing the power of Gemini’s machine learning algorithms, Reddit can better anticipate user intent, interpret complex queries, and deliver personalized search results tailored to each individual user.

    In practical terms, this integration means that Reddit users can now interact with Reddit Answers in a more conversational and intuitive manner. Whether they are seeking information, exploring new topics, or looking for recommendations, the AI-powered search tool powered by Google Gemini is poised to deliver faster, more accurate, and more personalized results.

    Furthermore, the integration of Google Gemini underscores Reddit’s commitment to leveraging cutting-edge technology to enhance the user experience on its platform. By staying at the forefront of AI and search innovation, Reddit is not only meeting the evolving needs of its users but also setting new standards for AI-powered search tools in the digital landscape.

    As Reddit and Google continue to collaborate and innovate in the AI space, we can expect to see further enhancements to Reddit Answers and other AI-powered features on the platform. The integration of Google Gemini is just the beginning of what promises to be an exciting journey towards more intelligent, intuitive, and user-centric search experiences on Reddit.

    In conclusion, Reddit’s integration of Google Gemini into Reddit Answers represents a significant milestone in the evolution of AI-powered search tools. By combining the capabilities of Gemini with the advanced AI infrastructure of Vertex AI, Reddit is poised to deliver a more sophisticated, personalized, and engaging search experience for its users. As the partnership between Reddit and Google continues to flourish, we can look forward to more innovations that will shape the future of search technology on the platform.

  • Why Clean Data Is the Foundation of Successful AI Systems

    In the fast-paced realm of artificial intelligence (AI), one fundamental truth reigns supreme: the quality of data is paramount to the success of AI systems. Recent research underscores this critical point, revealing that enterprises face potential annual losses amounting to a staggering $406 million due to subpar data quality, hindering the efficiency of their AI applications. Shockingly, these losses are projected to soar to an alarming $745 billion by 2025 if the issue remains unaddressed. It’s clear: clean data isn’t just a nice-to-have—it’s a non-negotiable technical prerequisite for developers and data engineers alike.

    The significance of clean data cannot be overstated. Imagine an AI system as a sophisticated machine—its algorithms and models are the gears and cogs that drive its functionality. However, without quality data as the lubricant, these components grind to a halt, impeding the system’s ability to perform optimally. Just as a car cannot run smoothly without proper maintenance and high-quality fuel, AI systems cannot function effectively without accurate, reliable data.

    At the core of every successful AI endeavor lies a foundation built upon clean, accurate, and relevant data. This foundational layer serves as the bedrock upon which AI models are trained and refined. Consider a scenario where an e-commerce platform utilizes AI to recommend products to its customers based on their browsing history. If the data feeding into the AI system is riddled with errors, inconsistencies, or biases, the recommendations generated will be flawed, leading to dissatisfied customers and lost revenue opportunities.

    Clean data acts as the lifeblood of AI systems, fueling them with the insights needed to make informed decisions and accurate predictions. By ensuring that data is free from duplications, inaccuracies, and incompleteness, organizations empower their AI systems to operate at peak performance levels, driving innovation, enhancing customer experiences, and unlocking new business opportunities.

    Moreover, clean data instills trust in AI outputs. In an era where data privacy and security concerns are at the forefront of discussions, ensuring the integrity and quality of data used in AI systems is paramount. Organizations that prioritize data cleanliness not only mitigate the risks of errors and biases but also build credibility with stakeholders, fostering a culture of transparency and accountability.

    To achieve and maintain clean data, organizations must implement robust data governance practices, establish data quality standards, and leverage advanced technologies such as data cleaning tools and algorithms. By proactively addressing data quality issues at every stage of the data lifecycle—from collection and storage to processing and analysis—businesses can fortify their AI systems against potential pitfalls and drive sustainable growth.

    In conclusion, the crucial role of clean data in the realm of AI cannot be understated. It serves as the linchpin that holds together the intricate machinery of AI systems, enabling them to function with precision, reliability, and integrity. As enterprises navigate the complexities of the digital landscape, investing in data quality initiatives is not just a strategic choice—it is an imperative that paves the way for AI-driven success in the data-driven era. Clean data isn’t just a foundation; it’s the cornerstone upon which the future of AI innovation rests.

  • AI Data Dilemma: Balancing Innovation with Ironclad Governance

    AI Data Dilemma: Balancing Innovation with Ironclad Governance

    Artificial Intelligence (AI) is revolutionizing industries with its ability to automate processes, personalize user experiences, and derive valuable insights from vast amounts of data. However, this innovation comes with a critical challenge: balancing the need for data-driven progress with stringent governance to protect privacy and ensure ethical use of information.

    In a landscape where data privacy regulations are increasingly stringent and AI models demand access to vast, diverse datasets, striking this balance is crucial. Organizations must navigate the delicate dance between leveraging data for innovation and upholding robust governance frameworks to safeguard against misuse.

    The Innovation Imperative

    Advancements in AI technologies rely heavily on access to high-quality, diverse datasets to train models effectively. Whether developing predictive analytics, natural language processing, or image recognition systems, AI algorithms thrive on the richness and breadth of data they can learn from.

    For example, in healthcare, AI algorithms trained on extensive patient data can help diagnose diseases more accurately and recommend personalized treatment plans. Similarly, in finance, AI-powered systems can analyze complex market trends and make rapid trading decisions based on vast amounts of historical data.

    The Governance Challenge

    Despite the immense potential AI offers, the misuse of data can have severe consequences, ranging from privacy breaches to perpetuating biases in decision-making processes. As a result, governing data usage and ensuring ethical AI practices have become top priorities for organizations and regulatory bodies alike.

    Maintaining ironclad governance involves implementing robust data protection measures, ensuring transparency in AI systems’ decision-making processes, and actively monitoring for any ethical lapses. By doing so, organizations not only mitigate risks but also build trust with consumers and stakeholders, fostering long-term relationships based on integrity and accountability.

    Striking the Balance

    To navigate the AI data dilemma successfully, organizations must adopt a proactive approach that harmonizes innovation with governance. This entails:

    • Data Governance Frameworks: Establishing clear policies and procedures for data collection, storage, and usage to ensure compliance with regulations and ethical standards.
    • Ethical AI Principles: Integrating ethical considerations into AI development processes, such as fairness, transparency, and accountability, to mitigate biases and promote responsible AI use.
    • Data Security Measures: Implementing robust cybersecurity protocols to safeguard sensitive information from unauthorized access or breaches, thereby preserving data integrity and trust.

    By embracing these practices, organizations can harness the full potential of AI while upholding ethical standards and regulatory requirements. This not only protects them from legal repercussions but also strengthens their reputation as responsible stewards of data and champions of innovation.

    In conclusion, the AI data dilemma underscores the critical need for organizations to find a delicate balance between driving innovation through AI and maintaining stringent governance practices. By prioritizing ethical considerations, data protection, and transparency, businesses can navigate this complex landscape successfully, reaping the benefits of AI while safeguarding against potential risks. In essence, the key lies in embracing innovation with a steadfast commitment to ironclad governance—a harmonious blend that paves the way for a sustainable and ethical AI-driven future.

  • Why Use a NoSQL Database for AI? There Are Many Great Reasons

    In the realm of cutting-edge technologies like Artificial Intelligence (AI), the choice of database plays a pivotal role in the success of projects. While traditional SQL databases have long been the default choice for many applications, the landscape is evolving, and NoSQL databases are gaining prominence in AI implementations. So, why should you consider using a NoSQL database for AI? Let’s delve into the compelling reasons driving this shift.

    Flexibility and Scalability

    One of the key advantages of NoSQL databases for AI applications is their flexibility and scalability. Unlike rigid SQL databases, NoSQL databases like MongoDB, Cassandra, or DynamoDB offer dynamic schema capabilities, allowing for seamless integration of diverse data types. This flexibility is crucial in AI projects where data formats can vary widely and evolve rapidly. Additionally, NoSQL databases are inherently designed for horizontal scaling, making them well-suited for handling the massive volumes of unstructured data typically involved in AI workloads.

    Speed and Performance

    When it comes to AI, speed is of the essence. NoSQL databases are optimized for performance, enabling quicker data retrieval and processing compared to traditional SQL databases. Their distributed architecture and ability to store and query large datasets in real-time make them ideal for AI applications requiring near-instantaneous responses. For tasks like real-time analytics, machine learning model training, or recommendation systems, the speed offered by NoSQL databases can significantly enhance overall system performance.

    Support for Unstructured Data

    AI algorithms thrive on unstructured data such as text, images, videos, or sensor data. NoSQL databases excel at handling these types of data, providing a natural fit for AI applications that rely on diverse data sources. By supporting unstructured data natively, NoSQL databases eliminate the need for complex data transformations or normalization processes, streamlining data pipelines and accelerating AI development cycles. This intrinsic support for unstructured data simplifies data ingestion, storage, and retrieval, empowering AI teams to focus on innovation rather than data wrangling.

    Distributed and Fault-Tolerant Architecture

    In the AI realm, where robustness and reliability are paramount, NoSQL databases shine with their distributed and fault-tolerant architecture. By distributing data across multiple nodes and ensuring automatic failover mechanisms, NoSQL databases enhance system resilience and minimize downtime. This distributed nature is particularly beneficial for AI applications deployed in cloud environments or edge devices, where data availability and consistency are critical. With NoSQL databases, AI systems can maintain high availability and performance even in the face of hardware failures or network disruptions.

    Integration with Big Data Ecosystem

    In today’s data-driven world, AI projects often rely on vast amounts of data from various sources within the Big Data ecosystem. NoSQL databases seamlessly integrate with popular Big Data technologies like Hadoop, Spark, or Kafka, enabling seamless data pipelines and analytics workflows. This interoperability simplifies data processing, transformation, and analysis, providing AI practitioners with a unified platform to harness the power of both AI and Big Data technologies. By leveraging NoSQL databases within the broader Big Data ecosystem, organizations can unlock new insights, drive innovation, and accelerate AI-driven decision-making.

    In conclusion, the adoption of NoSQL databases for AI applications is not just a trend but a strategic choice driven by tangible benefits. From flexibility and scalability to speed and performance, NoSQL databases offer a robust foundation for building AI systems that can handle the complexities of modern data environments. By embracing NoSQL databases in their AI journey, organizations can unlock new possibilities, drive innovation, and stay ahead in the competitive landscape of AI-driven technologies.

  • Three AI Superpowers: Classification AI vs Predictive AI vs Generative AI

    In the realm of artificial intelligence (AI), there exist three distinct superpowers that drive innovation and efficiency: Classification AI, Predictive AI, and Generative AI. Each of these AI “families” possesses unique capabilities that cater to specific needs and challenges within the tech landscape.

    Classification AI functions as a meticulous organizer, adept at sorting and labeling data into predefined categories. This AI powerhouse is what enables email platforms to sift through spam effortlessly or photo applications to identify familiar faces accurately. For instance, when you receive an email, and it magically lands in the correct folder without your intervention, you have Classification AI to thank for its precision.

    Predictive AI, on the other hand, acts as a virtual crystal ball, leveraging past data patterns to forecast future outcomes. Whether it’s predicting the weather for tomorrow or foreseeing a surge in consumer demand for a particular product, Predictive AI empowers businesses to make informed decisions and stay ahead of the curve. Imagine a scenario where an e-commerce platform predicts the next big trend in consumer behavior, allowing them to stock up on popular items in advance—this is the magic of Predictive AI in action.

    Generative AI embodies the essence of creativity within the AI landscape. This innovative force has the ability to generate entirely new outputs, such as lifelike images, emotive music compositions, or coherent text passages. By learning from existing patterns, Generative AI can produce content that mimics human-like creativity, opening up a world of possibilities for industries like art, music, and content creation. Picture an AI system that can compose symphonies or paint landscapes with a level of sophistication that rivals human artists—that’s the transformative power of Generative AI.

    These three AI superpowers work in harmony to revolutionize data organization, trend analysis, and creative content generation across various sectors. As AI technology continues to advance and evolve, the synergy between Classification AI, Predictive AI, and Generative AI will play a pivotal role in driving innovation and shaping the future of industries worldwide. Embracing the capabilities of these AI families is not just about leveraging cutting-edge technology—it’s about unlocking new possibilities and pushing the boundaries of what’s achievable in the digital age.

  • AI video startup Moonvalley raised a fresh $43M, SEC filing shows

    Moonvalley, an innovative AI video startup based in Los Angeles, has recently made headlines by securing a substantial $43 million in fresh funding through a successful venture capital round, as confirmed by an official SEC filing. This significant financial boost comes hot on the heels of Moonvalley’s recent introduction of Marey, their cutting-edge AI video-generating model, which has garnered attention and acclaim within the tech community.

    The fact that Moonvalley managed to attract such a sizable investment, especially during these challenging times, underscores the growing importance and potential of AI-driven technologies in the realm of video creation. With the backing of 11 undisclosed investors, Moonvalley is well-positioned to further develop and enhance its AI tools, paving the way for more groundbreaking solutions in the future.

    Moonvalley’s success story serves as a testament to the increasing demand for AI-powered innovations in various industries, including video production. By leveraging artificial intelligence, Moonvalley aims to revolutionize the way videos are created, offering efficient and cost-effective solutions that streamline the content creation process for professionals in the field.

    The launch of Marey, Moonvalley’s flagship AI video model, represents a significant milestone for the company and showcases its commitment to pushing the boundaries of technology in the video creation space. Marey’s capabilities are poised to empower content creators with advanced tools that can automate and enhance the video production workflow, ultimately leading to higher-quality output in less time.

    As Moonvalley continues to expand its offerings and solidify its presence in the market, the latest injection of funding will undoubtedly play a crucial role in fueling the company’s growth and innovation efforts. This infusion of capital not only validates Moonvalley’s vision and potential but also signifies investor confidence in the promising future of AI-driven video technologies.

    In a landscape where digital content creation is king, Moonvalley’s focus on AI tools for video production aligns perfectly with the evolving needs of creators and businesses alike. By harnessing the power of artificial intelligence, Moonvalley is at the forefront of a technological revolution that is reshaping the way videos are conceptualized, produced, and shared in today’s digital age.

    With a solid foundation of funding, a talented team of experts, and a groundbreaking AI video model in Marey, Moonvalley is well-equipped to lead the charge towards a future where intelligent automation and creativity converge seamlessly. As the company continues to push the boundaries of what’s possible in video creation, industry professionals can look forward to a new era of innovation and efficiency driven by AI-powered solutions like those offered by Moonvalley.

  • Shift-Right Testing: Smart Automation Through AI and Observability

    In the fast-paced realm of software development, traditional testing approaches have primarily focused on catching issues before the software hits the market—a practice known as shift-left testing. However, the accelerated speed of development facilitated by DevOps and CI/CD pipelines has unveiled a critical gap. Many issues that surface in real-world conditions, distinct from controlled pre-production environments, often remain undetected.

    Enter shift-right testing, a strategic pivot that addresses this very challenge. By enhancing testing strategies to include real-time production-like environments, organizations can fortify their quality assurance efforts. This shift emphasizes the significance of testing software in conditions that mirror live deployment scenarios, offering a more comprehensive evaluation of its performance and resilience.

    To bolster these efforts further, the integration of AI-driven observability emerges as a game-changer. AI technologies empower teams to gain deeper insights into how software functions in real-world settings. By leveraging AI-driven observability, QA automation engineers can proactively identify potential issues, predict system behavior, and optimize performance—all in real time.

    Imagine a scenario where an e-commerce platform experiences a sudden surge in traffic during a holiday sale. Shift-right testing coupled with AI-driven observability allows the team to anticipate and address potential bottlenecks before they impact user experience. By monitoring key metrics in real time and analyzing patterns with AI algorithms, teams can swiftly adapt to changing conditions and ensure seamless operation under varying loads.

    Moreover, AI-powered observability enables teams to delve into complex system interactions, uncover hidden dependencies, and detect anomalies that traditional testing may overlook. This granular level of insight not only enhances the overall quality of the software but also elevates the team’s ability to respond proactively to emerging challenges.

    In essence, the marriage of shift-right testing principles with AI-driven observability equips organizations with a proactive quality assurance approach. This dynamic duo enables teams to detect, diagnose, and resolve issues swiftly, fostering a culture of continuous improvement and innovation.

    As the digital landscape continues to evolve, embracing smart automation through AI and observability is no longer a luxury but a necessity. Organizations that invest in these cutting-edge practices position themselves not only to deliver superior software products but also to adapt swiftly to the ever-changing demands of the market.

    In conclusion, the convergence of shift-right testing, powered by AI-driven observability, marks a significant milestone in the evolution of quality assurance practices. By harnessing the collective power of automation, real-time testing, and AI insights, organizations can elevate their software testing capabilities to unprecedented levels. Embracing this paradigm shift is not just about staying ahead—it’s about redefining what’s possible in the realm of software quality and reliability.

  • Article: Bridging Modalities: Multimodal RAG for Advanced Information Retrieval

    Bridging Modalities: Enhancing AI with Multimodal RAG

    In the realm of Artificial Intelligence (AI), the integration of diverse modalities such as text, images, and audio has become a pivotal focus for advancing information retrieval systems. The concept of multi-model retrieval augmented generation (RAG) techniques is gaining significant traction for its ability to provide a more profound contextual understanding of data. Authors Suruchi Shah and Suraj Dharmapuram delve into the transformative potential of these techniques in their recent article, “Bridging Modalities: Multimodal RAG for Advanced Information Retrieval.”

    The core premise of multimodal RAG is to amalgamate various forms of data to enrich the learning and inference capabilities of AI systems. By leveraging text, images, audio, and potentially other modalities, these techniques enable AI models to comprehend information from multiple dimensions simultaneously. This holistic approach not only enhances the accuracy of information retrieval but also fosters a more nuanced understanding of complex datasets.

    One practical application highlighted in the article is the utilization of multimodal RAG in healthcare. Imagine an AI-powered system designed to assist medical professionals in diagnosing patients. By integrating text-based medical reports, visual data from scans or images, and even audio recordings of patient symptoms, the AI model can generate more comprehensive insights. This amalgamation of modalities empowers the system to provide more accurate diagnoses and personalized treatment recommendations, ultimately improving patient outcomes.

    The significance of multimodal RAG extends beyond healthcare into various domains such as e-commerce, education, and entertainment. In e-commerce, for instance, integrating text descriptions with product images and customer reviews can offer a richer shopping experience, leading to enhanced customer satisfaction and increased sales. Similarly, in education, multimodal learning platforms can cater to diverse learning styles by presenting information through text, visuals, and interactive elements, making the learning process more engaging and effective.

    One of the key advantages of multimodal RAG techniques is their ability to address the limitations of unimodal AI systems. Traditional AI models that rely on a single modality often face challenges in capturing the full context of a situation. For example, a text-only AI model analyzing a scene from a movie may struggle to interpret the emotions conveyed through facial expressions or background music. By incorporating multiple modalities, multimodal RAG enables AI systems to overcome these limitations and achieve a more holistic understanding of the content.

    In conclusion, the incorporation of multimodal RAG techniques represents a significant advancement in the field of AI and information retrieval. By bridging modalities and integrating diverse forms of data, these techniques open new avenues for enhancing contextual understanding, improving decision-making processes, and enriching user experiences across various applications. As technology continues to evolve, the adoption of multimodal RAG is poised to revolutionize the way AI systems interact with and interpret the world around us.

    In a rapidly evolving digital landscape, embracing the potential of multimodal RAG is not just a choice but a necessity for organizations looking to stay ahead of the curve. Whether in healthcare, e-commerce, education, or beyond, the ability to harness the power of multiple modalities for advanced information retrieval can unlock new possibilities and drive innovation. As we navigate the complexities of an increasingly interconnected world, the integration of multimodal RAG stands out as a beacon of progress, illuminating the path towards a more intelligent and insightful future.

  • How enterprise IT can protect itself from genAI unreliability

    In the fast-paced world of enterprise IT, the allure of generative AI (genAI) is undeniable. The promises of scalability, efficiency, and flexibility have captivated executives, driving them to explore the technology’s full potential. However, a shadow of unreliability looms over genAI, stemming from issues such as flawed deliverables, imperfect training data, and models that may overlook crucial specifics.

    Mayo Clinic has taken a proactive stance against unreliable genAI outputs by implementing innovative approaches. By combining algorithms like clustering using representatives (CURE) with large language models (LLMs) and vector databases, Mayo has enhanced data verification processes to ensure accuracy and consistency. This meticulous validation method serves as a blueprint for mitigating genAI unreliability within enterprises.

    To combat the inherent risks associated with genAI, enterprises have two primary strategies: human oversight and AI monitoring AI. While human intervention offers a sense of security, it may compromise the efficiency gains that genAI promises. On the other hand, AI self-monitoring presents a more futuristic yet daunting prospect, with organizations treading cautiously as they navigate this uncharted territory.

    The pivotal challenge lies in enhancing genAI reliability internally, as external solutions often fall short in delivering sustainable results. Integrating more transparency into genAI systems, such as requiring LLMs to acknowledge limitations or time constraints, can significantly boost output credibility. Additionally, establishing clear rules and protocols for AI agents, whether human or machine, is paramount in maintaining operational integrity.

    Senior management’s comprehension of genAI risks is crucial in shaping organizational risk tolerance levels and fostering a culture of accountability. As genAI continues to evolve, enterprises must adapt their environments to support these systems effectively. Managing the ecosystem around genAI, rather than just the model itself, is essential for long-term reliability and success.

    Despite the substantial investment required for deploying genAI models, the focus must shift towards ensuring reliable outputs to mitigate potential errors and liabilities. Embracing a proactive approach to managing genAI systems can safeguard enterprises from costly repercussions and reputational damage in the future.

    In conclusion, navigating the complexities of genAI reliability demands a strategic blend of technological sophistication, human oversight, and organizational adaptability. By embracing transparency, accountability, and continuous improvement, enterprises can harness the true potential of genAI while safeguarding against its inherent uncertainties. The choice to prioritize reliability is not just a matter of technology but a strategic decision that can shape the future trajectory of IT enterprises.