Aliquet morbi justo auctor cursus auctor aliquam. Neque elit blandit et quis tortor vel ut lectus morbi. Amet mus nunc rhoncus sit sagittis pellentesque eleifend lobortis commodo vestibulum hendrerit proin varius lorem ultrices quam velit sed consequat duis. Lectus condimentum maecenas adipiscing massa neque erat porttitor in adipiscing aliquam auctor aliquam eu phasellus egestas lectus hendrerit sit malesuada tincidunt quisque volutpat aliquet vitae lorem odio feugiat lectus sem purus.
Lorem ipsum dolor sit amet consectetur lobortis pellentesque sit ullamcorpe.
Mauris aliquet faucibus iaculis vitae ullamco consectetur praesent luctus.
Posuere enim mi pharetra neque proin condimentum maecenas adipiscing.
Posuere enim mi pharetra neque proin nibh dolor amet vitae feugiat.
The difficult of using AI to improve risk management
Viverra mi ut nulla eu mattis in purus. Habitant donec mauris id consectetur. Tempus consequat ornare dui tortor feugiat cursus. Pellentesque massa molestie phasellus enim lobortis pellentesque sit ullamcorper purus. Elementum ante nunc quam pulvinar. Volutpat nibh dolor amet vitae feugiat varius augue justo elit. Vitae amet curabitur in sagittis arcu montes tortor. In enim pulvinar pharetra sagittis fermentum. Ultricies non eu faucibus praesent tristique dolor tellus bibendum. Cursus bibendum nunc enim.
How to bring AI into managing risk
Mattis quisque amet pharetra nisl congue nulla orci. Nibh commodo maecenas adipiscing adipiscing. Blandit ut odio urna arcu quam eleifend donec neque. Augue nisl arcu malesuada interdum risus lectus sed. Pulvinar aliquam morbi arcu commodo. Accumsan elementum elit vitae pellentesque sit. Nibh elementum morbi feugiat amet aliquet. Ultrices duis lobortis mauris nibh pellentesque mattis est maecenas. Tellus pellentesque vivamus massa purus arcu sagittis. Viverra consectetur praesent luctus faucibus phasellus integer fermentum mattis donec.
Pros and cons of using AI to manage risks
Commodo velit viverra neque aliquet tincidunt feugiat. Amet proin cras pharetra mauris leo. In vitae mattis sit fermentum. Maecenas nullam egestas lorem tincidunt eleifend est felis tincidunt. Etiam dictum consectetur blandit tortor vitae. Eget integer tortor in mattis velit ante purus ante.
Vestibulum faucibus semper vitae imperdiet at eget sed diam ullamcorper vulputate.
Quam mi proin libero morbi viverra ultrices odio sem felis mattis etiam faucibus morbi.
Tincidunt ac eu aliquet turpis amet morbi at hendrerit donec pharetra tellus vel nec.
Sollicitudin egestas sit bibendum malesuada pulvinar sit aliquet turpis lacus ultricies.
“Lacus donec arcu amet diam vestibulum nunc nulla malesuada velit curabitur mauris tempus nunc curabitur dignig pharetra metus consequat.”
Benefits and opportunities for risk managers applying AI
Commodo velit viverra neque aliquet tincidunt feugiat. Amet proin cras pharetra mauris leo. In vitae mattis sit fermentum. Maecenas nullam egestas lorem tincidunt eleifend est felis tincidunt. Etiam dictum consectetur blandit tortor vitae. Eget integer tortor in mattis velit ante purus ante.
AI innovation never sleeps — and neither do we when it comes to noticing the advances in machine learning and artificial intelligence. This AI and Machine Learning Startups roundup list is to highlight disruptive startups, emerging technologies, and niche breakthroughs in our space. The list contains companies we talked to or interacted with throughout the month of November and December 2024.
Disclaimer: None of these entries are paid-for-advertisement or sponsored —this roundup is simply GMI Cloud's commentary on what excites us. The companies or projects named were not involved in creating this listicle, and this may be the first time they have heard about it.
Know a company that should make our list next month? Let us know at media@gmicloud.ai
Their tagline: Boost team productivity with business-grade intelligence across your data
What They Do: Storytell.ai is an enterprise-grade AI platform designed to transform raw data into actionable business insights for knowledge workers. The platform leverages its proprietary DistillAI algorithm, which helps separate signal from noise, and offers features like Story Tiles™ for organizing content and de-siloing knowledge across teams. Storytell's aim is to enhance collaboration and productivity within enterprises by making AI-driven tools accessible and effective for business professionals.
Why They Stand Out:
Colin: Storytell caught my attention because they solve a few real problems any organization faces: knowledge pooling, knowledge transfer, and knowledge synthesis. The tangible value for me (having played around with it) is that Storytell effectively functions as an inference engine that you don't really need to train. An org can just upload their company's content and knowledge into Storytell to result in a second brain that employees can leverage for inquiries and/or content creation.
Jonny: Makes it simpler and faster to use AI as a team using company specific information. Prompt engineering becomes easier because Storytell already has context about your company.
Concerns:
Colin: There are real data privacy concerns with a tool like Storytell. You're effectively uploading some very sensitive information and data to a third-party provider if you want to maximize what Storytell can do for you. While Storytell does have a strong commitment to security, any company using Storytell should have a deep discussion about information security, confidentiality, and privacy concerns when deciding what to upload to Storytell.
Jonny: It seems inevitable that there will be large competition for this use case. I imagine Google or Microsoft enabling a similar across company data rooms would make this easier and low barrier to entry for most companies.
Future Outlook:
If Storytell continues their current trajectory, we expect them to be an attractive option for organizations that want an institutionally relevant inference engine but don't have the resources to create and train one themselves.
Their tagline (translated): Visitors share their thoughts and questions in a confessional with a heavenly hologram, where words are only intended for four ears – but without confession. An artificial intelligence that reacts aesthetically as Jesus may create a sacred moment.
What They Do: This isn't a startup, but Peter’s Chapel in Lucerne, Switzerland has showcased an "experimental art installation" featuring AI Jesus in the confessional booth. Here's a video about it.
Why They Stand Out:
Colin: I mean… without getting into the concerns about whether this will draw in the wrath of God, my initial thought is that this will at least be interesting. I did a quick Googling and found out that in the early days of Church, confessions were public affairs. Then it became private confessions, and it stands to reason the practice can still advance from there. Whether the right move is to go AI, that's up to the individual, right? Would you use one?
Jonny: These niche use cases for AI are fascinating and show how AI is disrupting even the oldest of industries.
Concerns:
Colin: A Vice article noted: When the reporter entered the confessional, AI Jesus warned, “Do not disclose personal information under any circumstances. Use this service at your own risk. Press the button if you accept.” Confessions seem to be personal issues, which seems to be at odds with that warning.
Jonny: Does religion actually need AI? Maybe. When once there was a rush to build a mobile app for anything and everything, we might be seeing something similar with AI.
Future Outlook:
We're interested in seeing where religion takes AI, as organizations like Scientology, the Vatican, and more have definitely embraced technology over the years. Whether AI versions of people will be considered a spiritual reincarnation in the future remains to be seen.
Their tagline: The All-in-One Platform to Build and Deploy Vision AI
What They Do: Datature offers a platform designed to help companies build and deploy Vision AI applications. Its no-code MLOps tools enable users to transform expertise and visual data—such as videos, satellite images, and medical scans—into scalable AI solutions. The platform is used in various industries, including retail, GIS, manufacturing, agriculture, and healthcare, to derive actionable insights from visual data. Applications range from automating quality assurance to streamlining infrastructure inspection, aiming to enhance efficiency and improve outcomes.
Why They Stand Out:
Colin: There is probably a ton of use for an AI model to basically identify… well, anything you train it to identify. I wonder if the affordableness of platforms like these bring us closer to a post-CAPTCHA world. Remember a pre-CAPTCHA world? Well, I wonder what fresh hell awaits us when websites need new methods to defeat the bots.
Jonny: This powering some really incredible use cases – medical imaging, autonomous check-outs for retail, quality checks for manufacturing, best place to put solar panels on your house, and more. They have a great repository of resources and even offer a ‘Start for Free’ option for developers and researchers.
Concerns:
Colin: If I had to be nitpicky: full-customization vs. ease of deployment. "No-code" has been popular ever since people wanted to be able to create websites without knowing how to write HTML. However, we aren't certain if this concern truly applies to Datature, given that they're more of a platform for training your own desired Vision AI. There is still a lot to be said with having a low effort access to implement the latest best practices, so ease-of-use might be king.
Jonny: Honestly, nothing too concerning with this one. Would love to see more use cases.
Future Outlook:
If Datature can deliver on its promises, we expect this to go far. It's one thing to consciously go out and update yourself on best practices and then implement them, and Datature's value proposition of automating it for you cannot be ignored for deploying custom computer vision solutions.
Their tagline: An ideal solution for individuals and businesses looking to streamline their spreadsheet processes.
What They Do: GPTExcel is a tool that integrates GPT technology with Excel, enabling users to generate formulas and automate tasks within spreadsheets. It simplifies data manipulation and supports a range of Excel functionalities.
Why They Stand Out:
Colin: "How do I…?" is a common question we have when using tools with extensive features, but the specific function we're trying to accomplish is either esoteric or unfamiliar. Tools like GPTExcel definitely aid in that. I've definitely asked ChatGPT to generate an Excel formula for me before, so a purpose-generated tool is the logical next step. The fact that it has SQL Query generation, Regex generator, and even a Template generator just makes it better. The simple UI allows for easy onboarding and clearly gets the job done.
Jonny: This was an obvious use case for AI for many. GPTExcel has built what might be the best version of this tool that will clearly make the lives of many workers much easier and faster.
Concerns:
Colin: My concerns are mostly nitpicky and technical in nature: some text
What's going on under the hood? Given that it's a generated outcome, one must still view and test the results to ensure accuracy. Then there's the problem of user onboarding, given that it assumes a basic understanding of Excel spreadsheets. You still need to describe the formula you want to generate while trying to be as detailed as possible. Giving this tool to someone who isn't used to using Excel spreadsheets normally is an exercise in patience because they don't even know what is possible.
Jonny: Security – making sure that the service doesn’t have access to sensitive information like financial data.
Future Outlook:
We actually haven't had a chance to really do extensive testing on whether the paid version is significantly better than the free version, so we can't comment on that. The free tier is perfect for the casual user, and we do hope that tools like this increase in number and usability over time. Scope and feature creep is certainly a problem for some of the most common tools we have, and having these helpers will be useful.
Their tagline: The “easy button” for personalized sales outreach
What they do: Autobound.ai is a platform that helps B2B sales professionals generate hyper-personalized sales emails. It leverages AI to craft emails tailored to individual prospects by analyzing factors such as company news, buyer personas, LinkedIn profiles, and industry trends. Its goal is to automate and streamline the process of writing sales emails, enabling sellers to focus on higher-value activities like building relationships and closing deals.
Why it stands out:
Colin: If it delivers what it purports to do, then it's going to really streamline sales.
Jonny: Tools like these are really going to help certain types of workers do their job more effectively. In a business function where personalization has been paramount, this tool has the potential to turn your average CRM into a 10x CRM.
Concerns:
Colin: One concern with Autobound is the potential over-reliance on AI-generated content. While it creates highly personalized messaging, there's always a risk of it coming across as too formulaic or lacking a genuine human touch. Additionally, the tool's effectiveness may diminish if too many users in the same industry rely on similar templates, potentially creating messaging fatigue among prospects. Today, I'm already seeing a large amount of similar-looking and similar-sounding email outbounds in my spam folder. When everyone has AI-assisted outbounding, then those who use it are no longer as interesting. It's a weird Catch 22.
Jonny: I’m reminded of the infamous line from the move, The Incredibles, “If everyone is super, then no one is.” As these types of tools become more widely available and widely used, is personalization in emails even going to matter – or will the game evolve? Only time will tell.
Future Outlook: The future of Autobound looks promising, especially if it continues to evolve its AI to offer more dynamic and unique content tailored to different industries. Expanding beyond email to include voice, chat, or social media outreach could further solidify its position as a go-to tool for sales teams. Moreover, enhancing its analytics to provide actionable insights on what works best in specific scenarios could add tremendous value. But, don't lose your edge in writing customized, tailored emails when outbounding! Be prepared for any possible drastic changes in the email outbound meta, and it might come down to using these tools to have a starter draft. I still believe the final result of personalization must come from a human mind.
Their tagline: Advancing the frontier of AI systems with memory
What They Do: Letta is a cloud platform designed to create, deploy, and manage AI agents at scale. It provides tools for building production-ready applications powered by agent microservices, which use REST APIs for integration. The platform focuses on enhancing large language models (LLMs) with advanced memory capabilities through its MemGPT technology. This allows agents to maintain long-term memory, improving their reasoning and functionality over extended interactions.
Why They Stand Out:
Colin: I firmly believe context is everything, even if the relative event or piece of information is possibly old. I can totally see the value in giving LLMs a more advanced piece of memory.
Jonny: Advanced memory mixed with LLMs is a game changer. This is what is going to allow LLMs to become powerful databases for proprietary data. We like the open source model and white box systems that they Letta makes available.
Concerns:
Colin: Well, with advanced memory comes the concern about… event longevity? Memory longevity? Is there a term for this? What we're concerned with is: there's a reason people want the Right to be Forgotten, and this advanced memory will probably run head-on straight into this. I'm sure everyone would like to forget what they were like during their teenage years, and it would certainly be jarring for future AI to casually remind you about that time you wet your bed as a toddler, so you should probably buy the water-proof mattress.
Jonny: Concerns primarily revolve around data security and compliance. For services utilizing MemGPTs, there will likely need to be transparency surrounding what information is being stored and for how long.
Future Outlook:
We're really excited for AI to remember things over time! We just hope for companies and LLM models to remember that not everyone wants everything to be remembered, so please give users a way to say "Can you forget that, permanently?" (And then regret it in the future if it's somehow important again, ha.)
Their tagline: Bring your imagination to life with text, image, or video
What They Do: Sora by OpenAI is the latest paid platform for generating video. Whether you're organizing, editing, making shifts, trimming, or just asking it to create from scratch, Sora seems to be a very flexible tool for any project where the end result should be "a good video."
Why They Stand Out:
Colin: Being an OpenAI product certainly helps stand out. The generated videos can be hit or miss (but admittedly, with a higher chance of hitting). As a layperson, I can definitely see this being a useful tool for generating video for anyone who lacks the A/B-rolls, video editing skills, and video editing software. That being said, I'm not a video editor — I'm quite certain that the video editors and production people in the world are still significantly better than OpenAI.
Jonny: There is growing competition for text-to-video generators but what sets OpenAI and Sora to me, for now at least, is simply that people will use it because they already have access to ChatGPT. This is bringing text-to-video to the masses.
We're also pretty impressed with the feature that informs a user the exact cost of generation. Sora is NOT available for free as computing power is expensive. While most videos seem to fall under $1-2, there's still the problem of getting exactly what you want. There's much to be said about being economical with your tokens while trying to get a specific desired result. It might ultimately be better to hire a video editor to get a specific desired result while using Sora if you have an artistic or visual direction, but aren't looking for a prescriptive outcome.
Concerns:
Colin: As with anything here, there's a question of how it was trained, what materials it was trained on, and whether or not OpenAI truly secured the rights to use it. We're not lawyers here and the matter has yet to be resolved on whether these are appropriate uses, but just wanted to highlight this aspect. According to OpenAI's CTO, Sora was also trained on publicly available video, which may include YouTube, Facebook, and Instagram videos. How the original makers and actual copyright holders of those videos feel with regards to this usage is completely up to them.
Jonny: The breadth of this tool is still limited in terms of video quality, output size, and how many prompts a user can use per month. However, as the cost of inference goes down, there will be increasingly less limitations.
Future Outlook:
The existence of text-to-video generation has come a long way. Everyone remember that funny video of Will Smith eating spaghetti? That was less than two years ago! But there's just so many more hot topics to discuss with regards to text-to-video generation, such as deepfakes, illicit material, and more. That being said, a tool is rarely inherently good or bad and it is up to the user to determine whether it will be put to good use. The fact that this can be done has already opened the floodgates, and we really look forward to more of these in the future.
And that's it for this month's Spotlight list! We aim to do these each month, so if you know of an interesting product or solution that you want us to take a look at and comment on, don't hesitate to reach out to us at media@gmicloud.ai
Get started today
Give GMI Cloud a try and see for yourself if it's a good fit for AI needs.