Tech Republic

In today’s age of cybercrime, everybody should have a baseline of cybersecurity on their devices. Using a VPN is one of the easy things you can do to give yourself a little extra protection. And you don’t have to break the bank for a quality option, especially now that a lifetime subscription to FastestVPN has been reduced to just $24.97 at TechRepublic Academy. About FastestVPN FastestVPN offers protection for up to 15 devices at once and works on Windows, Mac, iOS, Android, smart TVs, and routers. The intuitive VPN gets its name for delivering 99.9% uptime and offering access to more than 600 high-speed servers all over the world. Each server utilizes military-grade 256-bit AES encryption and can allow you to bypass geographic restrictions to access the internet you want. The P2P-optimized servers also support high-speed downloading and streaming with complete anonymity. In addition to the servers, FastestVPN offers a NAT firewall for an extra layer of protection, malware protection to ensure the sites you visit are safe, and an ad blocker to improve your browsing experience. FastestVPN also offers a strict no-logging policy, ensuring your data stays hidden from everyone — even them. Smart Connect ensures you’re always on the optimal server for your browsing experience and you can choose between WireGuard®, IKEv2, and OpenVPN protocols. Find out why TenBestVPNs calls FastestVPN “one of the most promising VPN services in the market.” For a limited time, you can get a lifetime subscription to FastestVPN for up to 15 devices for one of the best prices you’ll find anywhere. Now you can sign up for just $24.97 (reg. $360) at TechRepublic Academy. Prices and availability are subject to change. source

Proton VPN Fast facts Our rating: 4.5 stars out of of 5Pricing: Starts at $4.99 per monthKey features Swiss-protected and open source VPN. Around 8,500 servers across 112 countries. Free version with unlimited data. In 2023, we found Proton VPN to be a well-rounded VPN with its unique, Swiss-based privacy and generous free tier. Based in Switzerland, Proton complies with Switzerland’s strong privacy laws — a plus for businesses and individuals concerned about government data requests. A year later, Proton VPN brings significant updates that I feel make one of the best VPNs available in the market. Chief among these is a larger server network, an updated third-party audit, and more affordable pricing. In this review, let’s take a look if Proton VPN is the right choice for you and your business in 2024. ESET PROTECT Advanced Employees per Company Size Micro (0-49), Small (50-249), Medium (250-999), Large (1,000-4,999), Enterprise (5,000+) Any Company Size Any Company Size Features Advanced Threat Defense, Full Disk Encryption , Modern Endpoint Protection, and more ManageEngine Log360 Employees per Company Size Micro (0-49), Small (50-249), Medium (250-999), Large (1,000-4,999), Enterprise (5,000+) Micro (0-49 Employees), Medium (250-999 Employees), Enterprise (5,000+ Employees), Large (1,000-4,999 Employees), Small (50-249 Employees) Micro, Medium, Enterprise, Large, Small Features Activity Monitoring, Blacklisting, Dashboard, and more Graylog Employees per Company Size Micro (0-49), Small (50-249), Medium (250-999), Large (1,000-4,999), Enterprise (5,000+) Medium (250-999 Employees), Enterprise (5,000+ Employees), Large (1,000-4,999 Employees) Medium, Enterprise, Large Features Activity Monitoring, Dashboard, Notifications Proton VPN pricing Proton VPN Free Proton VPN Plus Proton VPN Unlimited 1 year Free $4.99 per month $9.99 per month 2 years Free $4.49 per month $7.99 per month 1 month Free $9.99 per month $12.99 VPN speed Medium Fastest Fastest Main Features Servers in 5 countries 1 VPN connection Kill switch 6,500 – 8,500 servers in 110 countries 10 simultaneously connected devices Kill switch NetShield malware and ad blocker Streaming and torrenting available 30-day money-back guarantee All Proton VPN Plus features Includes Proton Mail Plus, Proton Drive Plus, and Proton Calendar When it comes to pricing, I found Proton VPN falls in the middle amongst its competition. At $4.99 per month for its one-year Plus plan, it’s more expensive than Private Internet Access’ equivalent one-year subscription at $3.33 per month. However, it’s also more affordable than ExpressVPN’s one-year plan priced at $8.33 per month. Compared to last year, Proton VPN is also now more affordable. In 2023, its one-year Proton VPN Plus subscription cost $5.99 per month. Today, it’s $4.99 per month. The two-year plan is also cheaper, now at $4.49 per month to 2023’s $4.99 per month. On the opposite side, its one-month subscription is slightly more expensive now at $9.99 per month, compared to last year’s $8.99 price tag. While these aren’t massive decreases in price, they are appreciated considering how the improvements implemented in Proton VPN in 2024 — something we’ll get into further down this review. If you plan on connecting multiple devices or need the fastest connection, I recommend going for Proton VPN Plus. It allows users to connect 10 devices simultaneously, has Proton’s fastest speeds at the lowest possible cost, and includes Proton’s full server suite. Those who also use VPNs for torrenting (P2P) and streaming should opt for the VPN Plus plan, as this is not available on Proton VPN’s free version. Lastly, I recommend Proton Unlimited only for users who want Proton’s other encryption-based services on top of the features you get with VPN Plus. Is Proton VPN actually free? Yes, Proton VPN has a fully free version alongside its two paid plans. It offers free access to servers from five countries, namely the United States, Netherlands, Japan, Romania, and Poland. ProtonVPN Free’s main interface. Image: Luis Millares With the free version, you get medium VPN speed, a kill switch, and unlimited data. Windows and Android users also get access to Proton’s split tunneling feature for free. As a free VPN, I appreciate Proton offering unlimited bandwidth, especially since other free VPNs usually impose a monthly or daily data cap. I think this implementation is better if you need a cost-effective way to simply change your IP address. However, there is a limitation I want to mention: Proton VPN’s Free version doesn’t let you choose which of the five free VPN locations you connect to. This means you will need to go through a bit of trial and error before you get connected to your desired country. As mentioned, you also won’t get access to the full 112-country server fleet available in its premium plan. Proton VPN’s restricted countries found in the premium version. Image: Luis Milllares While I find this to be a fair restriction, I can understand users that may want the control of being able to choose which location to connect to. If you’re one of these people, I suggest looking into Speedify VPN. It’s a decent service that has a free version VPN that allows manual country selection. Take note, however, that it imposes a 2-10 GB monthly data cap. SEE: Proton VPN Free vs Premium: Which Plan Is Best for You? (TechRepublic) In my view, the free version is best for users who are unsure about spending on a full VPN service or those who only need one VPN connection and basic functionality. More about Cloud Security Is Proton VPN safe? Yes, I consider Proton VPN to be a secure and safe VPN. It offers industry-standard protocols OpenVPN and IKEv2, as well as the newer WireGuard and Stealth protocols across the Windows, macOS, Linux, Android, and iOS versions. This gives users a mix of more established and well-tested security protocols and more recent offerings with newer technologies. Proton VPN’s tunneling protocols. Image: Luis Millares Proton VPN also operates under a strict no-logs policy, which means they don’t log user activity or metadata. This covers websites visited, session lengths, IP addresses, or any data that could be used to identify a user. It’s also an open source VPN, which means

Session Initiation Protocol (SIP) allows one device to find another and establish a connection over the internet. It’s a protocol that initiates a session — other protocols are responsible for transferring data. You can use SIP calling for voice calls, texting, video conferencing, and online messaging. Businesses can use SIP trunking to enable all of these channels for an entire office. For SIP calling, all you need is a stable internet connection and a device with speakers and a microphone. You may be familiar with Voice over Internet Protocol (VoIP), which relies on SIP and lets you make phone calls over the internet instead of the Public Switched Telephone Network (PSTN). Free VoIP phone services let you start SIP calling at no cost. To get started, you don’t have to know anything about SIP or VoIP. Just download an app to start making calls from a computer, tablet, or even a watch. The real question, though, is should you? 1 RingCentral Office Employees per Company Size Micro (0-49), Small (50-249), Medium (250-999), Large (1,000-4,999), Enterprise (5,000+) Medium (250-999 Employees), Enterprise (5,000+ Employees), Large (1,000-4,999 Employees) Medium, Enterprise, Large Features Hosted PBX, Managed PBX, Remote User Ability, and more What’s the catch with free SIP calling? For consumers, there’s not too much of a catch. You may have to endure some upsells in whatever SIP calling app you download, and there are bound to be inconveniences. It’s free. The people who make SIP apps usually want you to upgrade to a paid version, so anticipate a few snags. If you are just trying to stay in touch with family and friends, a free SIP calling is not bad at all. For business, I do not recommend free SIP calling. It’s a terrible idea. There are far better places to cut costs. Your business depends on your ability to communicate seamlessly with customers, prospects, partners, and vendors. A single-person business might be able to get by with free SIP calling — maybe — but for any larger organization, it falls well short of being a viable business communication solution. Here are the limitations you’ll face if you consider using free SIP calling. Lack of security: Calls may be vulnerable to interception and fraud. Unreliable quality: Dropped calls, poor audio, and latency are common issues. Limited support: Free services rarely provide timely or dedicated customer assistance. Scalability issues: Features like call routing or integrations are often unavailable. Hidden costs: Usage limits or necessary upgrades may make free plans costly. Intrusive ads: Some services include ads that undermine professionalism. On its own, each one of these reasons is enough to avoid free SIP calling for your business. Don’t you want rock solid VoIP security? Don’t you need call routing and other basic phone system administration features? I cannot, in good faith, recommend free SIP calling for business. You will be much better off with a cheap VoIP phone service, even if you are a solopreneur with the super basic phone system needs. Free SIP calling apps Most of these apps are simple and straightforward to use. For free SIP calling, you only need a device and a reliable internet connection. The app will turn your computer or tablet into a softphone capable of making calls over the internet. There are no special requirements other than choosing an app that suits you. You just need to download the app to your device and create an account. Most of these SIP calling providers have paid apps in addition to free ones. The free apps are fine for personal use, but I would definitely upgrade to a paid plan if you want to use SIP calling for your business. 1. Zoom Zoom is a popular VoIP application that allows customers to implement voice calls on any SIP-compatible device. Zoom phone is a feature-rich cloud-based phone system that provides on-demand scalability and connects to SIP trunks worldwide. Zoom Basic is its free plan. Here’s what you get: Video calls or meetings up to 40 minutes long Up to 100 participants per meeting Meeting captions Team chat End-to-end encryption There are a few more features, such as Zoom whiteboards, email, and a calendar client. You also get integrations with tools from Microsoft, Google, and popular CRM software like Salesforce and HubSpot. Zoom allows owners or administrators on iOS and Android mobile devices to record, forward, transfer, hold, and block calls. Upgrading to a paid Zoom plan gives you a lot more features, but the basic plan is fairly good compared to free plans from other providers. Check out our Zoom review to learn more about it. 2. OnSIP OnSIP is a hosted VoIP provider that allows you to make and receive calls to an SIP address with a free SIP account on any device (laptop, smartphone, or computer). OnSIP’s free SIP hosting lets you make voice or video calls on any standard SIP application. Therefore, OnSIP recognizes when you dial a SIP address number and automatically handles it as a free call. The free OnSIP account allows you to add up to 100 users with unlimited extensions. As a result, you have the leeway to easily build an entire network of SIP accounts for your team. The phone calls you make with your OnSIP network of users are handled within your local area network (LAN). Because it doesn’t go through the internet, you can make OnSIP calls to your LAN OnSIP network if your internet connectivity is down. Check out this OnSIP review for more information. 3. Zoiper Zoiper provides free softphones for non-commercial use alongside its paid versions. It offers mobile downloads for Android and iOS and Linux, Mac, and Windows versions for the desktop. In addition to SIP, Zoiper also supports the Inter-Asterisk eXchange (IAX) protocol. Zoiper is used by businesses, service providers, and operators, such as VoIP integrators, mobile operators, and call centers, that need softphones independent of their service providers. For your convenience, Zoiper aggregates and integrates your contacts from email services such as Outlook,

NVIDIA revealed various infrastructure, hardware, and resources for scientific research and enterprise at the International Conference for High Performance Computing, Networking, Storage, and Analysis, held Nov. 17 to Nov. 22 in Atlanta. Key among these announcements was the upcoming general availability of the H200 NVL AI accelerator. The newest Hopper chip is coming in December NVIDIA announced at a media briefing on Nov. 14 that platforms built with the H200 NVL PCIe GPU will be available in December 2024. Enterprise customers can refer to an Enterprise Reference Architecture for the H200 NVL. Purchasing the new GPU at an enterprise scale will come with a five-year subscription for the NVIDIA AI Enterprise service. Dion Harris, NVIDIA’s director of accelerated computing, said at the briefing that the H200 NVL is ideal for data centers with lower power — under 20kW — and air-cooled accelerator rack designs. The H200 NVL is targeted toward low-power HPC and AI workloads. Image: NVIDIA “Companies can fine-tune LLMs within a few hours” with the upcoming GPU, Harris said. H200 NVL shows a 1.5x memory increase and 1.2x bandwidth increase over NVIDIA H100 NVL, the company said. Dell Technologies, Hewlett Packard Enterprise, Lenovo, and Supermicro will support the new PCIe GPU. It will also appear in platforms from Aivres, ASRock Rack, GIGABYTE, Inventec, MSI, Pegatron, QCT, Wistron, and Wiwynn. SEE: Companies like Apple are working hard to create a workforce of chip makers. Grace Blackwell chip rollout proceeding Harris also emphasized that partners and vendors have the NV GB200 NVL4 (Grace Blackwell) chip in hand. “The rollout of Blackwell is proceeding smoothly,” he said. Blackwell chips are sold out through the next year. Must-read big data coverage Unveiling the Next Phase of Real-Time Omniverse Simulations In manufacturing, NVIDIA introduced the Omniverse Blueprint for Real-Time CAE Digital Twins, now in early access. This new reference pipeline shows how researchers or organizations can accelerate simulations and real-time visualizations, including real-time virtual wind tunnel testing. Built on NVIDIA NIM AI microservices, Omniverse Blueprint for Real-Time CAE Digital Twins lets simulations that generally take weeks or months be performed in real time. This capability will be on display at SC’24, where Luminary Cloud will show how it can be leveraged in a fluid dynamics simulation. “We built Omniverse so that everything can have a digital twin,” Jensen Huang, founder and CEO of NVIDIA, said in a press release. “By integrating NVIDIA Omniverse Blueprint with Ansys software, we’re enabling our customers to tackle increasingly complex and detailed simulations more quickly and accurately,” said Ajei Gopal, president and CEO of Ansys, in the same press release. CUDA-X library updates accelerate scientific research NVIDIA’s CUDA-X libraries help accelerate the real-time simulations. These libraries are also receiving updates targeting scientific research, including changes to CUDA-Q and the release of a new version of cuPyNumeric. Dynamics simulation functionality will be included in CUDA-Q, NVIDIA’s development platform for building quantum computers. The goal is to perform quantum simulations in practical times — such as an hour instead of a year. Google works with NVIDIA to build representations of their qubits using CUDA-Q, “bringing them closer to the goal of achieving practical, large-scale quantum computing,” Harris said. NVIDIA also announced the latest cuPyNumeric version, the accelerated scientific research computing library. Designed for scientific settings that often use NumPy programs and run on a CPU-only node, cuPyNumeric lets those projects scale to thousands of GPUs with minimal code changes. It is currently being used in select research institutions. source

Microsoft Ignite 2024, held Nov. 19 – 22 in Chicago, featured nearly 100 announcements and software updates, including an AI feature in Teams that can translate speech and replicate an individual employee’s voice. This year’s overarching theme was expanding generative AI’s summarization and rewriting capabilities to address more niche use cases. AI translator agent can replicate your voice in Teams Microsoft is going all-in on AI “agents” in an effort to further abstract the workings of large language models. Ideally for Microsoft, this would make those models more capable of autonomous, sequential actions. Microsoft Copilot Studio will enable autonomous agents soon, with this feature now in preview. Other major announcements around agents at Ignite 2024 include: Agents in SharePoint, available now, can answer questions about files, folders, or sites. A Teams Facilitator agent, which takes notes and summarizes meetings, is now in preview. An Interpreter agent in Teams, which can replicate a user’s voice in up to nine languages for real-time translation, will be in preview in early 2025. Copilot Studio will soon include images and voice content Copilot Studio and Azure AI Foundry will be more closely linked, with a Microsoft Agent SDK available for both developers to create custom agents. The Agent SDK, available now in preview, can draw from Azure’s large AI model catalog. Soon, Copilot Studio users can experiment with multimodality in the agents they build. Image upload and analysis are now in preview, and voice is in private preview. (Private preview in Azure is invite-only.) SEE: As of this Patch Tuesday, Copilot PC users can remap the AI button. More must-read AI coverage Azure AI Foundry offers new capabilities Microsoft announced a new way to access AI in the Azure AI Foundry, a hub for AI models and solutions. Azure AI Foundry is accessible in preview through either an SDK or online portal and interoperates with GitHub and Visual Studio. Both offer slightly different options: the SDK helps admins and app developers evaluate AI models and test apps before deploying them. The portal replaces the former Azure AI Studio and offers management of AI apps, models, and services. Redmond also announced a service allowing developers to make and deploy AI agents. Azure AI Agent Service will be in preview in December. In other Azure news: Azure AI will offer an AI scorecard containing “model cards, model versions, content safety filter configurations, and evaluation metrics.” The goal of AI reports is to help development teams complete audits and compliance reports. Risk and safety evaluations for image content will flag harmful content generated by AI. The go is to help with data-driven assessments of how often such content might appear. Cloud professionals can now map out AI adoption in the Cloud Adoption Framework. AI workload guidance has also been added to the Azure Well-Architected Framework. Azure AI Content Understanding, now in preview, is a streamlined workflow for turning unstructured text, images, and video into a corpus of data. It packages together templates, grounding mechanisms, and confidence scores to help developers and enterprises release AI solutions. Developers can run AI workloads on serverless GPUs in Azure Container Apps, which are currently in preview. Security Exposure Management tool enters general availability In the security world, Microsoft introduced a new service: Security Exposure Management. This service works in concert with Extended Detection and Response, Microsoft Defender XDR in Microsoft Security. It provides insight into an organization’s security posture. Now available to all users, Security Exposure Management had previously been in preview only. Deploying Security Exposure Management helps raise awareness of an organization’s attack surface, analyzes attack paths, and aggregates security posture data into “Unified Exposure Insights.” These insights, in turn, are translated into metrics to measure security initiatives. In other security news, Security Copilot in Intune for workflow management will be expanded. Security Copilot will now be able to weigh in on a broader range of scenarios across Intune, the Intune Suite, and Windows Autopatch. Intune admins can now add iOS, Android, macOS, and Linux devices to their hardware inventory. Microsoft announced several similar updates for Microsoft Purview, but we won’t cover them exhaustively here. See Microsoft for more information about the 90-plus announcements from the Ignite event. Copilot+ PCs add flexibility to Windows Search Starting in early 2025, Search will look different in File Explorer, Windows Search, and Settings. Users can search for documents and photos using natural language, and an internet connection isn’t required. Microsoft encourages business customers to show off prompts to coworkers In today’s AI infrastructure, companies tend to produce tools that make adopting their AI products easier. For example, Copilot Analytics, set to expand its features in early 2025, offers insights into adoption and usage trends to measure business impact. Additionally, Microsoft is gamifying AI adoption by enabling easy prompt sharing and introducing a prompt leaderboard. Both features will be part of the Copilot Prompt Gallery — formerly Copilot Labs. Prompt sharing will roll out with Microsoft 365 Copilot later in 2024, while an organization-wide “new and trending prompts” list is expected in early 2025. One new feature, generative AI controls in Microsoft 356 Copilot, encourages caution with generative AI. The controls will notify users if prompts or responses in Microsoft 365 Copilot contain information that might be sensitive or put an organization at risk. Admins can configure Microsoft Purview Data Loss protection individually to match the sensitivity of their organization’s data. Is AI ROI panning out? In a recent survey, Slack revealed that AI adoption might be slowing down. Why? Some people feel the technology hasn’t lived up to the hype, that AI use is considered “lazy,” or that they don’t know how to use AI. However, in a November 2024 study, Microsoft reported a 10x ROI increase among “top” business leaders using generative AI. Most AI use cases observed by Microsoft were focused on productivity. TechRepublic covered Microsoft Ignite remotely. source

As cybercrime has grown over the past few years, the demand for ethical hackers has reached a new height. These security specialists are paid to break into corporate networks and systems to expose vulnerabilities, thereby finding what needs to be patched up and protected. It’s a lucrative career, and whether you’re looking for ways to protect your business or add another revenue stream, learning ethical hacking could be a smart move. Another smart move would be to get started with The All-in-One Super-Sized Ethical Hacking Bundle. This 18-course bundle gives you a beginner-to-intermediate education in ethical hacking and it’s on sale now for just $34.97 through December 8. What’s included This massive bundle includes courses from some of the web’s top instructors, including Gabriel Avramescu (4.4/5-star instructor rating), Atul Tiwari (4.2/5-star rating) and Amit Huddar. Through these courses, you’ll get a beginner-friendly education to ethical hacking, learning some of today’s most important tools and technologies that help ethical hackers do their jobs. When you’re starting out, you’ll cover the basics of ethical hacking, learning how to hack websites and systems before delving into more complicated projects. Through the hands-on courses, you’ll get real-world experience of how to fend off and protect against real cyberattacks, giving you the confidence to do it in a corporate environment, too. From Burp Suite and BitNinja to Kali Linux, Metasploit and more, you’ll discover the necessary tools to succeed as an ethical hacker. Protect your business from an onslaught of cybercrime and potential losses. Right now, you can get The All-in-One Super-Sized Ethical Hacking Bundle for just $34.97 for a limited time. Prices and availability are subject to change. source

Savvy businesses have discovered a secret weapon in the battle against software subscription fees: Lifetime licenses. Yes, your days of monthly or yearly subscriptions could very well be over if you grab this new PDF tool. SwifDoo PDF Pro doesn’t require a subscription to edit, convert, sign, or manipulate your business’ PDFs. Instead of those painful $20 monthly fees, you’d only have to pay $34.99 once to own this lifetime PDF software for Windows (reg. $129). Goodbye expensive Adobe, hello affordable SwifDoo We understand it might be hard to part with a tool that’s done your business well, but it’s time to cut back on expenses. PDFs are a crucial part of operations, but a basic tool shouldn’t cost you $240 yearly—especially when SwifDoo can do the same things. Here are just some of this affordable PDF software’s features: Add and edit text to update contracts, resumes, or reports. Crop and adjust images to tweak pictures or enhance a presentation. Create PDFs from various formats like Microsoft Office documents, CAD drawings, or images. Scan paper files to digital to digitize documents or notes. Translate documents when working with international clients. Like any good PDF tool, SwifDoo also has OCR. This means you can search a document for text without reading through the entire thing manually. You might search for a name, number, or quote and easily copy-paste it from the PDF. Aren’t you glad it’s not Adobe? Check out now to get this Adobe Acrobat alternative for only $34.99, compared to $20 monthly for the rest of your business days. Get SwifDoo PDF Pro here for an unbeatable price. Prices and availability subject to change. source

Retrieval Augmented Generation (RAG) has become essential for IT leaders and enterprises looking to implement generative AI. By using a large language model (LLM) and RAG, enterprises can ground an LLM in enterprise data, improving the accuracy of outputs. But how does RAG work? What are the use cases for RAG? And are there any real alternatives? TechRepublic sat down with Davor Bonaci, chief technology officer and executive vice president at database and AI company DataStax, to find out how RAG is being leveraged in the market during the rollout of generative AI in 2024 and what he sees as the technology’s next step in 2025. What is Retrieval Augmented Generation? RAG is a technique that improves the relevance and accuracy of generative AI LLM model outputs by adding extended or augmented context from an enterprise. It allows IT leaders to use generative AI LLMs for enterprise use cases. Bonaci explained that while LLMs have “basically been trained on all the information available on the internet,” up to a certain cut-off date, depending on the model, their language and general knowledge strengths are offset by significant and well-known problems, such as AI hallucinations. SEE: Zetaris on why federated data lakes are the future for powering AI “If you want to use it in an enterprise setting, you must ground it in enterprise data. Otherwise, you get a lot of hallucinations,” he said. “With RAG, instead of just asking the LLM to produce something, you say, ‘I want you to produce something, but please consider these things that I know to be accurate.’” How does RAG work in an enterprise setting? RAG gives an LLM reference to an enterprise information set, such as a knowledge base, a database, or a document set. For instance, DataStax’s main product is its vector database, Astra DB, which enterprises are using to support the building of AI applications in enterprises. In practice, a query input given by a user would go through a retrieval step — a vector search — identifying the most relevant documents or pieces of information from a pre-defined knowledge source. This could include enterprise documents, academic papers, or FAQs. The retrieved information is then fed into the generative model as additional context alongside the original query, allowing the model to ground its response in real-world, up-to-date, or domain-specific knowledge. This grounding reduces the risk of hallucinations that could be deal breakers for an enterprise. AI resources from TechRepublic Premium How much does RAG improve the output of generative AI models? The difference between using generative AI with and without RAG is “night and day,” Bonaci said. For an enterprise, the propensity for an LLM to hallucinate essentially means they are “unusable” or only for very limited use cases. The RAG technique is what opens the door to generative AI for enterprises. “At the end of the day, they [LLMs] have knowledge from seeing things on the internet,” Bonaci explained. “But if you ask a question that is kind of out of the left field, they’re going to give you a very confident answer that may … be completely wrong.” SEE: Generative AI has become a source of costly mistakes for enterprises Bonaci noted that RAG techniques can boost the accuracy of LLM outputs to over 90% for non-reasoning tasks, depending on the models and the benchmarks used. For complex reasoning tasks, they are more likely to deliver between 70-80% accuracy using RAG techniques. What are some RAG use cases? RAG is used across several typical generative AI use cases for organisations, including: Automation Using LLMs augmented with RAG, enterprises can automate repeatable tasks. A common use case for automation is customer support, where the system can be empowered to search documentation, provide answers, and take actions like canceling a ticket or making a purchase. Personalisation RAG can be leveraged to synthesize and summarise large amounts of information. Bonaci gave the example of customer reviews, which can be summarised in a personalised way that is relevant to the user’s context, such as their location, past purchases, or travel preferences. Search RAG can be applied to improve search results in an enterprise, making them more relevant and context-specific. Bonaci noted how RAG helps streaming service users find movies or content relevant to their location or interests, even if the search terms don’t exactly match available content. How can knowledge graphs be used with RAG? Using knowledge graphs with RAG is an “advanced version” of basic RAG. Bonaci explained that while a vector search in basic RAG identifies similarities in a vector database — making it well-suited for general knowledge and natural human language — it has limitations for certain enterprise use cases. In a scenario where a mobile phone company offers multiple-tiered plans with varying inclusions, a customer inquiry — such as whether international roaming is included — would require the AI to decide. A knowledge graph can help organise information to help it figure out what applies. SEE: Digital maturity key to success in AI for cybersecurity “The problem is the content in those plan documents are conflicting with each other,” Bonaci said. “So the system doesn’t know which one is true. So you could use a knowledge graph to help you organise and relate information correctly, to help you resolve these conflicts.” Are there any alternatives to RAG for enterprises? The main alternative to RAG is fine-tuning a generative AI model. With fine-tuning, instead of using enterprise data as a prompt, data is fed into the model itself to create an influenced data set to prime the model for use in a way that can leverage that enterprise data. Bonaci said that, to date, RAG has been the method widely agreed upon in the industry as the most effective way to make generative AI relevant for an enterprise. “We do see people fine-tuning models, but it just solves a small niche of problems, and so it has not been widely accepted as a solution,” he said. source

TL;DR: Turbocharge your programming, particularly your collaborations, with Microsoft Visual Studio Professional 2022 for Windows — it’s on sale for just $27.97 through November 21. Time is money, so boosting your productivity can have a significant impact on your bottom line, such as when you can collaborate on programming seamlessly and efficiently across platforms with your teams. That is exactly what Microsoft Visual Studio Professional 2022 for Windows allows you to do, and it’s on sale right now and through November 21, at just $27.97. Code faster Less typing means faster coding. With IntelliCode, Visual Studio can complete a line or an entire block of code for you. It can even list the next best options to choose from, all of which help you code more accurately, as well as more rapidly. Meanwhile, CodeLens offers a comprehensive view of your codebase, displaying authors, recent changes, tests and more so you can make more informed decisions. Collaborate Live Share offers personalized coding sessions with real-time collaboration, including access controls and customized editor settings to speed up editing and debugging cycles. Even more importantly, it ensures that everyone’s code remains consistent. Multi-language programming Best of all, Visual Studio allows you to program across platforms and languages. Create cross-platform desktop and mobile apps. Edit ASP.NET pages as they’re running using the web designer view. Use Blazor to design responsive web user interfaces in C# or build, test and debug C++ and .NET apps in Linux. You can even employ hot reload capabilities across both of them. Why Microsoft Visual Studio Basically, you can turbocharge your productivity by using advanced tools and built-in integrations to craft high-quality code. Then you can deliver the most innovative apps, no matter how challenging your development workflows are. It’s easy to see why Visual Studio has a perfect 5-star rating on Microsoft Choice Software. It’s always good for business when you can learn new skills, but it’s every bit as important to have the tools that can make your work better, easier and more efficient. Get Microsoft Visual Studio Professional 2022 for Windows today while it’s on sale for just $27.97. Sales end November 21 at 23:59 Pacific Time. Prices and availability subject to change. source

Artificial intelligence is becoming crucial to how large retail organisations manage and optimise supply chains. From predicting seasonal demand in goods to automating inventory ordering, AI is helping supply chain management system vendors gain new efficiencies for their clients. In 2022, McKinsey reported that supply chain management was the top area where businesses reported AI-related cost reductions. At the time, large consumer packaged goods companies saw a 20% reduction in inventory, a 10% decrease in supply chain costs, and revenue increases of up to 4%. AI for supply chains has only improved since 2022 and is accelerating with generative AI. A more recent report from McKinsey found that supply chain management was the function where businesses most commonly reported meaningful revenue increases of more than 5% due to investments in AI. Machine learning has done the grunt work of optimising supply chains Laurence Brenig-Jones, vice president of product strategy at supply chain management and planning software provider RELEX Solutions, told TechRepublic the “number crunching” power of machine learning has been the dominant AI technology force used in supply chains to date. “I think what we are seeing is there is a huge improvement in accuracy and automation [from machine learning capabilities] that can lead to very significant benefits in product availability, reduction in working capital, and if you’re a grocer, then a reduction in spoilage or wastage,” he said. There are several use cases for which machine learning has been deployed in supply chains. More must-read AI coverage Demand forecasting Predicting product demand is key in supply chain management. Brenig-Jones said this is “incredibly difficult” because it can involve predicting demand for a specific product, at a specific location, on a specific day or time of day — often up to 180 days or more in advance across an entire operation. Over the last five years, machine learning algorithms have replaced previously used time series algorithms for this task. According to ERP vendor Oracle, AI can now use internal data such as sales pipelines and external signals like market trends, economic outlooks, and seasonal sales for forecasting. Automated inventory Demand forecasting helps organisations optimise and automate inventory ordering. Though this includes ensuring sufficient stock is available to meet demand, retailers must also balance other factors, such as excessive working capital with too much stock, food spoilage, or capacity breaches. Brenig-Jones said many optimization algorithms, with their ability to learn from the past through machine learning, can solve this complex problem and efficiently fulfill demand for the organisation’s supply chain, balancing all involved factors. Logistics optimisation Machine learning is also embedded in logistics networks. According to Oracle, logistics companies use machine learning algorithms to “train models that optimise and manage the delivery routes by which components move along the supply chain,” ensuring more timely deliveries of goods. SEE: Supply chain job openings point to a lack of automation and innovation In one example, courier company UPS uses its dynamic road-integrated optimisation and navigation platform, ORION, to show drivers the most efficient route for deliveries and pickups on more than 66,000 roads in the U.S., Canada, and Europe, saving significant mileage and fuel costs annually. The growing role of generative AI in supply chain management Experts believe generative AI will become increasingly important in supply chain management and planning. Through natural language queries, the future will likely see an expanded role for generative AI. Richer natural-language interactions Retailers will likely have much richer and more analytical natural-language interactions with their supply chain and retail planning data in the future. This could involve asking questions about the supply chain plans, what has happened in the past, or where there are opportunities to do better. “You could ask: ‘What were my top five reasons for out-of-stocks last week?’ And it could tell you: ‘Number one was poor inventory accuracy in your stores, and these stores in particular. Number two was you had one big supply failure, and it caused this impact on your sales’, Brenig-Jones said. Forward-looking recommendations Generative AI in supply chain management platforms could offer forward-looking recommendations for large retailers through natural language interactions. For example, a platform could advise an organisation on what to do next week to ensure everything is set up to hit its targets. SEE: Splunk urges Australian organisations to secure LLMs “It might say: We recommend that you change this part of your configuration, or we recommend you go and talk to this supplier because there’s a risk based on our understanding of what happened last time.’ So it would be forward-looking and interacting in a natural language format,” Brenig-Jones said. Becoming an AI ‘superuser’ A further phase in the introduction of generative AI, and something RELEX is pursuing within its platform, is to turn AI into a “super user.” Like system users who are “real gurus in how the system is configured,” AI could become self-adaptive, helping organisations improve their systems over time. “It would say: ‘I’ve come up with a better configuration for your solution based on what I’m seeing,’” Brenig-Jones explained. “So you would get into this kind of ability for the solution to self-adapt on the go. That is the direction we’re heading, and we’re working with our customers to understand how that would work best for them as well.” source