devopsy.kinsta.cloud

Tag: AI tools

  • Breaking Boundaries: Integrating Generative AI for Developers

    Breaking Boundaries: Integrating Generative AI for Developers

    Many organizations are experiencing deficits in technical skills and gaps in productivity that ultimately harm their bottom line. To overcome those obstacles and drive growth, businesses can use generative AI tools like ChatGPT and Bard to handle large volumes of data and increase the speed and output of basic tasks without draining valuable human capital. Already, businesses are finding that saving time is one of the most valuable benefits generative AI can deliver. However, there must be oversight of deployments to avoid recurring pitfalls, as the technology contains just as much risk as it does opportunity.

    Generative AI provides developers with helpful task automation while simultaneously creating new challenges. Security, bias and IP questions have DevOps teams scratching their heads at how to best ethically and operationally implement generative AI. Developers must embrace these challenges to remain competitive in an increasingly automated and technologically-driven world.

    Maintaining Confidentiality of Company Code

    Generative AI’s ability to understand and solve code-related issues makes it an ideal tool to help software developers optimize code for faster feature development and quicker releases. By identifying where it can contribute to tasks, like legacy code conversion or automation of processes, companies can free up valuable developer time for building the tools that can grow the business or drive new initiatives. An added benefit is that development teams can automate the coding of new applications or systems by updating old code or re-using existing examples.

    ChatGPT, for example, can also turbocharge development time by scouring swathes of information on GitHub and other forums to find examples of whatever a coder is trying to build. A 2022 study published in the Proceedings of the Association for Computing Machinery on Programming Languages (PACMPL) found that developers could code faster and work more efficiently thanks to generative AI. Tedious searches in tech forums and manual copy-and-paste procedures are replaced by simply asking the right questions (via prompts).

    However, generative AI tools are not closed systems. As users enter data into ChatGPT, it is saved and aggregated to continuously train the language model, opening the door for potential leaks of sensitive company information. There is a serious, potential risk of employees using AI tools as sensitive data already makes up 11% of what employees paste into ChatGPT.

    As a third-party system, AI tools absorb every query, comment and question without any privacy guardrails for users. If an AI bot’s security is compromised, businesses can unintentionally hand over data to bad actors. Even without a hacker’s involvement, sharing confidential customer or partner information may violate contractual agreements and create long legal battles. Developer teams must remain conscious of AI’s security risks.

    Avoiding Bias Creeping Into Results

    It is of utmost importance that businesses make reliable, fair and traceable decisions based on the latest facts. Currently, ChatGPT-4 hasn’t reached the maturity level necessary to be trusted by most companies without significant verification measures. Already, testers have persuaded ChatGPT to make statements that are either inaccurate, biased or both.

    One key concern in generative AI is the ingrained and accidental bias in answers due to the human-generated content used to train large language models. ChatGPT works by searching for textual content relating to the question and recombining it for a distillation. In an almost democratic way, the majority opinion will dominate. For instance, if most people believe that the Swedish are hard drinkers, ChatGPT will say so. While the technology provides remarkably accurate responses in an alarmingly convincing way, it’s important to remember that it’s not always the most reliable or balanced source that wins.

    Furthermore, OpenAI, the creator of ChatGPT, has explicitly stated that its goal is to create a system where queries can be pre-processed before they are fed to the neural net. Answers can then be post-processed before being shown to users. But this safeguard lacks clarity around its objectives and opens the door for ideological or political manipulation.

    Companies considering implementing AI tools must carefully examine their potential applications to ensure they avoid scenarios where bias can result in insights and output that could end up harming the company and its reputation.

    Minimizing Developer Function

    Enterprises are looking for tools to integrate generative AI into existing business processes to enhance numerous functions and deliver greater value to their customers. While the technology still isn’t capable of replacing human developers responsible for updating and maintaining largescale software, updates to generative AI tools and integrations are occurring at a rapid pace, leaving businesses to determine which programs are most viable for developer function.

    As the economy continues to tumble closer to a recession, companies seek ways to minimize costs. And those looking to use generative AI to work with the source code and technical documentation the company uses to run its own operations—rather than the open source code tools like ChatGPT provide—face budgetary decisions. Implementing generative AI programs into the business model can be extremely costly if done wrong, leaving executives to weigh the benefits of adopting the tech versus the cost of trying to hire a full team of developers.

    Connecting the Enterprise to Avoid AI Pitfalls

    For businesses looking to build the connected experiences their employees, partners and customers desire, generative AI is a great resource to accomplish this. However, it’s just one piece of the puzzle that is a modern enterprise. These tools can only deliver smart answers if they get smart questions and businesses connect their internal processes and applications to enable machine learning on accurate and holistic datasets.

    Although AI poses many benefits for developers, such as increased efficiency, accuracy and the ability to handle large amounts of data, they must be wary of the potential security, ethical and budgetary concerns that impact employees, customers and the enterprise as a whole.

    However, once the concerns are overcome or managed, the power of large language models will be leveraged in all kinds of technology and business use cases. Certainly a truly transformational shift and a step change towards a world where technology and software play an unprecedented role in business and society.

  • Revolutionizing the Nine Pillars of DevOps With AI-Engineered Tools

    Revolutionizing the Nine Pillars of DevOps With AI-Engineered Tools

    In my book “Engineering DevOps,” published a few years ago, I classified DevOps into “nine pillars of DevOps best practices” – a comprehensive framework that covered the full scope of DevOps practices. Each of these pillars represents a vital area of DevOps implementation and, together, they ensure a holistic approach to integrating development and operations for effective collaboration and continuous delivery.

    The current generation of AI-engineered tools is now being applied to each of the DevOps pillars and has the potential to significantly increase the efficiency, reliability and speed of the software delivery process. By automating repetitive tasks, improving decision-making with predictive insights and facilitating proactive issue resolution, AI can drive a higher level of quality, security and performance in the software development life cycle. Furthermore, AI’s capacity to learn and adapt over time means continuous improvement, fostering a truly data-driven, learning-oriented DevOps culture.

    AI Applications for the Nine Pillars of DevOps

    The following paragraphs briefly explain how the current generation of AI-engineered large language model (LLM) tools are being applied or could be practically applied to improve each of the nine pillars of DevOps.

    Leadership Practices: Leadership is vital to drive cultural changes, set vision and goals, encourage collaboration and ensure resources are allocated properly. Strong leadership fosters a successful DevOps environment by empowering teams and supporting innovation. AI can assist leaders in decision-making by analyzing large datasets to identify trends and predict outcomes, providing valuable insights to guide strategic planning.

    Collaborative Culture Practices: DevOps thrives in a culture of openness, transparency and shared responsibility. It’s about breaking down the silos that can exist between different teams (like development, operations and QA) and promoting effective communication and collaboration. AI-powered tools can improve collaboration through smart recommendations, fostering more effective communication and knowledge sharing.

    Design-for-DevOps Practices: This involves designing software in a way that supports the DevOps model. This can include aspects like microservices architecture, modular design and considering operability and deployability from the earliest stages of design. AI can be used in the software design phase to automatically suggest improvements in code quality or to identify potential performance issues early.

    Continuous Integration Practices: This is about developers frequently merging code changes into a central repository. After that, automated builds and tests are run. The main objectives are to detect problems early and reduce integration issues. AI can analyze patterns in past build failures to predict potential issues with new builds, leading to proactive error mitigation.

    Continuous Testing Practices: Continuous testing involves automating tests to provide instant feedback on business risks and allowing for faster, more reliable releases. This approach enhances the quality of the software and reduces the time it takes to resolve defects. AI can be used to generate and optimize test cases automatically, reducing manual effort and speeding up the testing process.

    Elastic Infrastructure Practices: Elastic infrastructure involves automated provisioning and management of computing resources, often leveraging cloud-based solutions for scalability and resilience. This includes practices like infrastructure-as-code (IaC). AI can help manage cloud resources by predicting future needs based on usage trends and automatically adjusting resources to optimize cost and performance.

    Continuous Delivery Practices: This is an approach where code changes are automatically built, tested and prepared for release to production. It aims to make releases less disruptive and more frequent, increasing speed and efficiency. AI can help predict the potential impact of new releases on the system, aiding in decision-making related to deployment.

    Continuous Security Practices: Also known as DevSecOps, this involves integrating security practices into the DevOps workflow. The goal is to introduce security considerations early and often in the life cycle rather than leaving them as an afterthought. AI can be used in anomaly detection to identify potential security threats and automate security policy enforcement.

    Continuous Monitoring Practices: Monitoring systems and application performance in real-time allows teams to detect and react to problems quickly, often before they impact users. It’s crucial for maintaining application health and informing future improvement efforts. AI can analyze logs and metrics to predict potential system failures or performance degradation, allowing for proactive maintenance and issue resolution.

    Pitfalls and Challenges

    Applying AI to DevOps is a complex process with certain challenges. Here are some potential pitfalls, along with ways to address them:

    Data Quality and Availability: AI relies heavily on data for training and making accurate predictions. If the data is inaccurate, incomplete or biased, it can lead to misleading outcomes. Ensure you’re collecting high-quality, relevant and unbiased data for training AI models. Regular data audits can help maintain data integrity.

    Over-Reliance on Automation: While automation is beneficial, over-reliance on it without human oversight can result in overlooked errors or system failures. Maintain a balance between automation and human intervention. Implement periodic manual checks and reviews to validate AI-generated results and actions.

    Security and Privacy Concerns: As AI tools are more integrated into the DevOps pipeline, they may become targets for security threats. Also, the handling of sensitive data used for AI training can raise privacy issues. Implement strong security practices, conduct regular security audits and consider privacy requirements during the data collection process.

    Black Box Problem: AI, especially complex models like neural networks, often operate as a “black box,” meaning it’s difficult to understand why they made a certain decision. This lack of transparency can make troubleshooting difficult. Consider using explainable AI (XAI) techniques, which aim to create more understandable AI systems, providing transparency into how decisions are made.

    Skills Gap: The application of AI in DevOps requires a certain level of expertise and understanding of both domains, which can be hard to find. Invest in training and upskilling existing staff, or consider bringing in external expertise through hiring or consultancy. Sharing knowledge and encouraging continuous learning can also help bridge this gap.

    Summary

    AI-engineered tools can enhance all of the nine pillars of DevOps by increasing efficiency, speeding up processes and making predictions to aid decision-making. However, implementing AI in DevOps comes with challenges such as ensuring data quality, maintaining a balance between automation and human intervention, addressing security and privacy concerns, managing the “black box” problem of AI and overcoming skills gaps. Each of these pitfalls can be addressed through proactive strategies like maintaining data integrity, implementing manual checks, bolstering security practices, using explainable AI techniques and investing in training and skills development.

  • Using Automation to go All-Digital, Fast

    Using Automation to go All-Digital, Fast

    Before the pandemic, digital transformation was a five-year plan for many companies — or in some cases, a distant vision that kept getting kicked down the road. Now, companies in every industry face an urgent imperative to adapt to an increasingly online-first world: If you don’t go digital, fast, you’re simply out of the game.

    Increasing productivity has been critical for companies as they make this shift. But too often, speed is hindered by time-consuming and repetitive processes that companies can and should easily automate. By automating mundane tasks, employees and businesses can focus on what really matters — increasing scale in today’s work-from-anywhere world, and improving experiences for both employees and customers.

    The Benefits of Automation

    Fundamentally, automation is about digitizing existing business processes to streamline results, and it will be a key element in helping businesses survive and get back to growth. Until now, many companies have focused on applying these tools to the front office: sales, marketing, commerce and supply chains. With most employees now working remotely, it’s harder than ever for team members to do their jobs efficiently. Automating the back office and improving the employee experience is key for businesses to achieve the speed they need to thrive right now.

    Here are three key advantages of bringing automation to the back office:

    1. It improves productivity.In virtually every industry and role, employees are still spending hours every week on mundane and labor-intensive tasks, such as data entry. Automation can speed up these repetitive tasks and streamline lengthy processes, freeing employees up to focus on the “human” aspects of their jobs, like interacting with customers, planning ahead, making decisions or learning new skills. By taking over routine tasks and allowing employees to concentrate on what matters most, automation makes workers more efficient and productive.
    2. It increases scale.There’s only so much an employee can do manually. The time-consuming processes that worked yesterday can’t scale to meet the demands of an all-digital environment.  Automating repetitive tasks means that employees can achieve more in the same amount of time, allowing them to scale their workloads. With time-consuming tasks being taken care of in the background, workers can spend their time on more meaningful work and customer relationships.
    3. It enhances employee experience.Nobody enjoys entering numbers into a spreadsheet. An often overlooked aspect of automation is that it leaves employees more satisfied and engaged with their companies and their jobs. Having fewer tedious tasks on their plates gives employees the chance to tap the skills that machines aren’t as good at, like creativity, communication and decision-making. This is especially beneficial at a time when most employees are juggling more responsibilities at work and at home. Making a more valuable contribution and more meaningful connections is not only rewarding for employees, but also translates to higher satisfaction among the customers they interact with.

    Real-World Examples

    Not everything can be automated, but there’s a lot of room to take the drudgery out of most roles. Take onboarding, for example. Traditionally, when a new hire started, everything happened in person: They attended an orientation, picked up their laptops at the tech bar, met with benefit providers — all of which required orchestration on the part of human resources. With automation, all of this manual coordination can be replicated with an intelligent workflow capable of guiding employees through an onboarding process customized to their specific role and location.

    In another example, HR departments are facing an unprecedented volume of queries about policies around caring for children and other dependents. Using low-code development, companies can easily set up automated, online bots to handle frequently asked questions and guide employees through processes needed to enroll in new services. Automating these queries with bots not only makes the employees themselves more productive, but it also delivers better experiences for both the HR team and the employees seeking answers.

    Automation is Here to Stay

    In today’s world, there is no alternative to a digital business model. Automation is the recipe that will help companies achieve the speed they need to survive now and the edge they need to thrive in the future.

    And there is no turning back — automation is here to stay, regardless of the trajectory of the pandemic. The process of change is never easy, but done right, automation not only leads to a more agile workforce, but also to a happier one.

  • 6 Common Pitfalls to Avoid While Building Out Your AI Tools

    6 Common Pitfalls to Avoid While Building Out Your AI Tools

    Enterprise-level companies, with thousands of employees across the globe, can have their AI framework replicated by B2B companies of nearly any size. In fact, even without access to a ton of proprietary data and immense computational power, you can build out effective AI tools.

    To begin with, you can introduce publicly available datasets from movie review websites, e-commerce review sites and elsewhere to train your system. When doing so, be sure to double check the license details of such publicly available corpuses, as some will not allow commercial usage. Then, during the preprocessing stage—after removing swear words, jargon and other irrelevant elements from your corpus—you can add noise and slightly alter your existing data to create even more relevant data points in your corpus. 

    After this, you can build a semi-supervised learning algorithm through techniques such as tri-training. In tri-training, we essentially train three different neural networks over binary classification data. Once two models agree with one another, the data is introduced to the third network and then re-trained until all three models are in agreement. However, before your AI tools are up and running, it’s important to be mindful of the following common pitfalls.  

    Overfitting AI Models

    It may seem counterintuitive, but you really do not want your AI model to be too accurate. If your cross validation score is anything above 100, then you might need to reconsider deploying the model, since it might be prone to overfitting. If your training data is overfit, one small change can render the entire model useless. Your model should be accurate. However, it cannot be too accurate, as it needs to be adaptive. 

    Class Imbalance

    When training your AI tools, it sometimes makes sense to include a disproportionate amount of your target variable. For example, perhaps you’re training your computer vision AI tool to recognize planks of wood in a warehouse. Even if planks of wood only account for 1% of your total inventory, you’ll want to oversample the minority class. In this case, you’ll likely want 60% of your sample to be planks of woods. You really want to amplify this number to ensure you have the right class balance. 

    Data Leakage

    Another common misstep is the failure to capture data leakage, which can occur in any number of ways. A common instance occurs when an element used to test the model slips into the training data. This generally causes your model to appear to be performing well. However, its predictions might not be as good when deployed in production. To rectify this, it is vital that you keep track of any changes to your AI tools. In addition to your organization’s code reviews and configuration change reviews, you should have data set reviews to ensure your AI engineering is being done properly. Any parameter change within your AI tools needs to be fastidiously documented. 

    Concept Drift

    All AI tools are dependent upon the shape of the data that comes into your system. For example, perhaps your AI model was trained for response times being between 1-100 seconds, and all of a sudden the response times are between 1-1,000 seconds. In this case, your model will likely fail to capture the new data points. It’s important to be aware of such an occurrence, so that you can retrain your model accordingly. 

    There are many techniques that can be used to identify instances of concept drift. Rather than waiting for feedback from your customers, you can set up anomaly detection in a way that it identifies concept drift on incoming data and automatically updates your AI model.

    Using Personal Data

    With the CCPA and GDPR in place and a federally mandated U.S. data privacy law on the horizon, organizations are wary of using their customers’ data to train their AI tools. It’s prudent to train your AI system with anonymized data; this way, you won’t set yourself up for problems down the road. 

    Failing to Ensure All Your AI Decisions Are Explainable

    Most AI-enabled decisions are enacted upon by teams rather than individuals. Hence, it’s vital that your AI model’s recommended courses of action come with explanations. Additionally, these explanations should have confidence intervals, making it easier to interpret results. 

    During root cause analysis and remediation, it’s often unclear which team is ultimately responsible for the issue at hand. If the AI’s suggested course of action comes with an explanation, this helps immensely during the remediation process. For example, if the starter event is a network build-up, the issue likely lies with the engineering team. Alternatively, if the starter event is a network configuration change, the network team is the likely culprit. By closely monitoring your AI tools’ performance and ensuring that all AI decisions are explainable, you can quickly solve issues and course correct.

  • Ensure Your AI is Working for Your Data Team

    Ensure Your AI is Working for Your Data Team

    AI tools hold great promise for DataOps. They can help teams identify the best apps to migrate, identify optimum cloud infrastructure and even validate decisions. But how can an IT team ensure the AI tools it uses are actually advancing its DataOps goals?

    A hard pill to swallow is that even the best-designed algorithmic decision-making and AI tools are prone to errors. Fortunately, this is unlike errors in software development, which are largely caused by development. AI errors emerge from poorly defined project goals and the organization’s tendency to treat the technology as if it’s magic. Without solid requirements, model monitoring and regular check-ins with reality, it is almost impossible for AI to deliver.

    This is why audits must be a mandate when introducing any AI tool that greatly influences business decisions. However, before the audit is conducted, ensure that the following points are met:

    Ensure the AI Tool is ‘Auditable’

    This step ensures that organizations bring in the right AI tools to their arsenal.

    An AI tool is auditable when it is:

    • Transparent: When AI tools are used to enhance the decisions a business makes, stakeholders must be able to peer into the system to ensure the legitimacy of the algorithm decision. Ensure the AI tool you choose has a logging system that makes it transparent as to how the system is making a decision, what factors most influence it and how confident it is in that decision. This allows decision-makers to trust in the system, as well as creates a feedback loop for the developers to take in real-world data and enhance and fine-tune the system to your needs.
    • Built with best practices: Just as security engineers follow best practices when securing a network, AI models and tools built with best practices in mind fare better when in use. As a general rule of thumb, all AI models must keep detailed access and process logs, incorporate failback mechanisms and generate regular health reports.

    Set Up Markers for Success

    Don’t approach an AI audit as though it’s a sandbox question. If teams are clear as to what an AI tool must achieve and how to measure its success, only then will an audit be able to judge the performance of the tool.

    It is helpful to set up markers that describe success. This could be the number of cases the AI model has rightly predicted, the time it takes, the number of edge cases it is able to handle and the number of times the AI decision collided with a human decision.

    The Audit

    Once the AI tool has been integrated into the organization and is in active use for a defined period, audits must occur. They help you answer the following questions:

    • Is the algorithm suitably transparent to end-users?
    • Is it likely to be used in a socially acceptable way?
    • Might it produce undesirable psychological effects or inadvertently exploit natural human frailties?
    • Is the algorithm being used for a deceptive purpose?
    • Is there evidence of internal bias or incompetence in its design?
    • Is it adequately reporting how it arrives at its recommendations and indicating its level of confidence?

    (HBR: “Why we need to audit algorithms”)

    Assess the Algorithm’s Fitness for its Purpose

    From a use-case perspective, AI algorithms in modern software applications tend to fall into two main categories. One is offering insights that help humans make decisions—this is what AI-driven analytics is all about. The other is allowing software applications to undertake automated actions entirely on their own, as AIOps tools increasingly do.

    A basic step in auditing your algorithm is to determine which type of use case it is supposed to support, then evaluate how well it achieves it. This is important to do regularly, because AI use cases can drift as applications evolve and the original purpose of an algorithm might change over time. (It’s common for algorithms to be implemented originally as a way to offer insights, but for applications to be updated with the expectation that the algorithms can support automated decision-making.)

    Compare and Contrast Algorithmic Decisions Against Human Decisions

    Though some of the inner workings of an AI model are a black box to even data scientists, the result of those inner workings is not. An AI audit must set up a process and team that regularly and actively monitors the decisions AI makes and compares them against decisions that a human would make.

    This is useful in identifying early on whether the AI tool has internalized biases. With an AI system, input greatly affects learning. For example, if a tool used to forecast purchase orders is incorporated by the organization during peak sales season, the tool will assume that this is the organization’s baseline performance. This causes erratic predictions in the long run.

    Active Experimentation

    Though an audit should uncover any bugs, biases or security concerns of an AI tool, it also could be used to test system variables. An AI tool is generally built on assumptions of the problem borders and trends. However, assumptions are very fragile in reality. The baselines of reality could change, causing the AI to produce questionable results.

    An audit also must be able to revisit AI tools with new assumptions, annually or biannually, to uncover any new parameters that could help improve the accuracy of the tool.

    Interviewing Users of the System

    Ultimately, an AI tool is put into place to either reduce the workload of humans or enhance it. Daily users have vast knowledge and perspective into not just how the tool works but also its limitations and quirks.

    Conclusion

    More often that not, the mystique of an AI model is used as an excuse for erratic results and poor performance. By building a robust and transparent audit system, your organizations can ensure your AI tools are actually working for your teams—and if not, it can at least uncover rare scenarios and edge cases and help mitigate them.

    This sponsored article was written on behalf of Unravel.

    — Swaathi Kakarla