Datadog Inc., one of the top dogs in the application monitoring software business, today announced the launch of new large language model observability features that aim to help customers troubleshoot problems with LLM-based artificial intelligence applications.

The new features were announced alongside the launch of its own generative AI assistant, which helps dig up useful insights from observability data.

Datadog is a provider of application monitoring and analytics tools that are used by developers and information technology teams to assess the health of their apps, plus the infrastructure they run on. The platform is especially popular with DevOps teams, which are usually composed of developers and information technology staff.

DevOps is a practice that involves building cloud-native applications and frequently updating them, using teams of application developers and IT staff. Using Datadogs platform, DevOps teams can keep a lid on any problems that those frequent updates might cause and ensure the health of their applications.

The company clearly believes the same approach can be useful for generative AI applications and the LLMs that power them. Pointing out the obvious, Datadog notes generative AI is rapidly becoming ubiquitous across the enterprise as every company scrambles to jump on the hottest technology trend in years. As they do so, theres a growing need to monitor the behavior of the LLM models that power generative AI applications.

At the same time, the tech stacks that support these models are also new, with companies implementing things like vector databases for the first time. Meanwhile, experts have been vocal of the danger of leaving LLM models just to do their own thing, without any monitoring in place, pointing to risks such as unpredictable behavior, AI hallucinations where they fabricate responses and bad customer experiences.

Datadog Vice President of ProductMichael Gerstenhaber told SiliconANGLE that the new LLM observability tool provides a way for machine learning engineers and application developers to monitor how their models are performing on a continuous basis. That will enable them to be optimized on the fly to ensure their performance and accuracy, he said.

It works by analyzing request prompts and responses to detect and resolve model drift and hallucinations. At the same time, it can help to identify opportunities to fine-tune models and ensure a better experience for end users.

Datadog isnt the first company to introduce observability tools for LLMs, butGerstenhaber said his companys goes much further than previous offerings.

Abig differentiator is that we not only monitor the usage metrics for the OpenAI models, we provide insights into how the model itself is performing, he said. In doing so, our LLM monitoring enables efficient tracking of performance, identifying drift and establishing vital correlations and context to effectively and swiftly address any performance degradation and drifts. We do this while also providing a unified observability platform, and this combination is unique in the industry.

Gerstenhaber also highlighted its versatility, saying the tool can integrate with AI platforms including Nvidia AI Enterprise, OpenAI and Amazon Bedrock, to name just a few.

The second aspect of todays announcement is Bits AI, a new generative AI assistant available now in beta that helps customers to derive insights from their observability data and resolve application problems faster, the company said.

Gerstenhaber explained that, even with its observability data, it can take a great deal of time to sift through it all and determine the root cause of application issues. He said Bits AI helps by scanning the customers observability data and other sources of information, such as collaboration platforms. That enables it to answer questions quickly, provide recommendations and even build automated remediations for application problems.

Once a problem is identified, Bits AI helps coordinate response by assembling on-call teams in Slack and keeping all stakeholders informed with automated status updates,Gerstenhaber said. It can surface institutional knowledge from runbooks and recommend Datadog Workflows to reduce the amount of time it takes to remediate. If its a problem at the code-level, it offers concise explanation of an error, suggested code fix and a unit test to validate the fix.

When asked how Bits AI differs from similar generative AI assistants launched by rivals such as New Relic Inc. and Splunk Inc. earlier this year,Gerstenhaber said its all about the level of data it has access too. As such, its ability to join Datadogs wealth of observability data with institutional knowledge from customers enables Bits AI to assist users in almost any kind of troubleshooting scenario. We are differentiated not only in the breadth of products that integrate with the generative interface, but also our domain-specific responses, he said.

THANK YOU

View original post here:

Datadog announces LLM observability tools and its first generative ... - SiliconANGLE News

The advancements in Language Model (LM) technologies have revolutionised natural language processing and text generation. Among these, Large Language Models (LLMs) like GPT-4, Bard, Claude etc. have garnered significant attention for their impressive capabilities. However, the deployment of LLMs in business settings raises concerns regarding privacy and data security,andleaked informationisattheorderoftheday. In this comprehensive article, we will delve into the negative consequences of using personal information in LLM prompts for businesses and the urgent precautions they must take to safeguard user data.

Over the course of 2023, businesses have increasingly tapped into the untapped potential that Large Language Models have. From professional experience, common use cases involve the integration of personal information into LLM prompts. This poses a severe risk of privacy breaches,aswellasbiasedoutputsstemmingfromuncheckeddatasets. Businesses also often use customer data to personalise content generation, such as chatbot responses or customer support interactions. However, including sensitive user information in prompts could lead to unintended exposure, jeopardizing customer privacy and undermining trust.

For instance, if a chatbot accidentally generates a response containing personal identifiers like names, addresses, or contact details, it could inadvertently divulge sensitive information to unauthorized individuals. Such privacy breaches can lead to legal consequences, financial losses, and damage to a business's reputation.

Businesses globally are subject to data protection laws and regulations that govern the collection, storage, and usage of personal data. By utilising personal information in LLM prompts without appropriate consent and security measures, businesses risk non-compliance with data protection regulations like GDPR (General

View post:

The Danger of Utilising Personal Information in LLM Prompts for ... - Medium

Amazon Web Services, the king of renting cloud servers, is facing an unusually large amount of pressure. Its growth and enviable profit margins have been dropping, Microsoft and Google have moved fasteror opened their walletsto capture more business from artificial intelligence developers (TBD on whether it will amount to much), and Nvidia is propping up more cloud-provider startups than we can keep track of.

Its no wonder AWS CEO Adam Selipsky last week came out swinging in an interview in response to widespread perceptions his company is behind in the generative AI race.

With Amazon reporting second quarter earnings Thursday, the company undoubtedly is trying to get ahead of any heat coming its way from analysts wondering whats up with AWS and AI. The company dropped some positive news Wednesday last week at a New York summit for developers. AWS servers powered by the newest specialized chips for AI, Nvidia H100 graphics processing units, are now generally available to customers, though only from its North Virginia and Oregon data center hubs.

See original here:

The AWS Empire Strikes Back; A Week of LLM Jailbreaks - The Information

This year, Salesforce has been making significant strides in the field of generative AI with the introduction of their large language models (LLMs). These LLMs, including their own Salesforce LLM, have proven to be highly effective in various use cases such as sales, service, marketing, and analytics.

Salesforces LLM has outperformed expectations in testing and pilot programs, producing accurate results when asked to provide answers. This puts Salesforce on the forefront of AI technology in the customer relationship management (CRM) space.

Other providers of AI models include Vertex AI, Amazon Sagemaker, OpenAI, and Claude, among others. These models can be trained to produce optimal results for organizations leveraging them. However, effective training requires large amounts of data, which can be stored in data lakes provided by companies like Snowflake, Databricks, Google BigQuery, and Amazon Redshift.

Salesforces LLM leverages Data Cloud, allowing flexibility in working with GenAI and Salesforce data. With Data Cloud, organizations can enjoy pre-wiring to Salesforce objects, reducing implementation time and improving data quality. Salesforces three annual releases also ensure a continuous stream of new and improved capabilities.

Salesforce has built an open and extensible platform, allowing integration with other platforms to bring in data from different sources alongside CRM data. This approach, known as Bring Your Own Model, enables organizations to use multiple providers/models simultaneously, preventing any potential conflict among machine learning teams.

Salesforces investments in GenAI technology organizations, demonstrated by their AI sub-fund, further solidify their commitment to advancing AI in the CRM space. These investments include market leaders like Cohere, Anthropic, and You.com.

While no LLM is 100% accurate, Salesforce has implemented intentional friction, ensuring that generative AI outputs are not automatically applied to users workflows without human intervention. Salesforce professionals working with GenAI have the freedom to use their preferred models and are provided with upskilling resources to effectively implement GenAI in their organizations.

The future of GenAI in CRM looks promising, with Salesforce constantly exploring new use cases and enhancements for their LLM technology. This creates opportunities for Salesforce professionals to advance their careers in the AI space.

Go here to read the rest:

Salesforce's LLM and the Future of GenAI in CRM - Fagen wasanni

Photo by Xu Haiwei on Unsplash

I am going to write this first blog to share my learning of Large Language Models (LLM), Generative AI, Langchain, and related concepts. Since I am new to the above topics, I will add a few concepts in 1 blog.

Large language models (LLMs) are the subset of artificial intelligence (AI) that are trained on huge datasets of written articles, blogs, texts, and code. This helps them to create written content, and images, and answer questions asked by humans. These are more efficient than the traditional Google search we have been using for quite some time.

Though new LLMs are still added nearly daily by developers and researchers all over the globe, they have earned quite a reputation for performing the tasks below:

Generative AI is the branch of AI that can create AI-powered products for generating texts, images, music, emails, and other forms of media.

Generative AI is based on very large machine-learning models that are pre-trained on massive data. These models then learn the statistical relationships between different elements of the dataset to generate new content.

LLM and Generative AI though are fresh technologies in the market, they are already powering a lot of AI-based products and there are startups that are raising billions.

For example, LLMs are being used to create chatbots that can easily have natural conversations with humans. These chatbots could be used to provide customer service, psychological therapy, act as financial or any specific domain advisor, or just can be trained to act as a friend.

Generative AI is also being used to create realistic images, paintings, stories, short to long articles, blogs, etc. These are creative enough to trick humans and will keep getting better with time.

With time these technologies will keep getting better and let humans work on more complicated tasks thus eliminating the need for mundane repetitive tasks.

This marks the end of the blog. Stay tuned, and look out for more python related articles, EDA, machine learning, deep learning, Computer Vision, ChatGPT, and NLP use cases, and different projects. Also, give me your own suggestions and I will write articles on them. Follow me and say hi.

If you like my articles please do consider contributing to ko-fi to help me upskill and contribute more to the community.

Github: https://github.com/abhigyan631

Read more:

LLM and Generative AI: The new era | by Abhinaba Banerjee | Aug ... - DataDrivenInvestor

Google is determined to boost Google Assistant by integrating LLM (large language model) technology, according to a leaked internal memo. The restructuring within the company aims to explore the possibilities of enhancing Google Assistant with advanced features. The memo emphasizes Googles commitment to Assistant, as it recognizes the significance of conversational technology in improving peoples lives.

Although the memo does not provide specific details, it suggests that the initial focus of this enhancement will be on mobile devices. It is expected that Android users will soon be able to enjoy LLM-powered features, such as web page summarization.

The leaked memo does not mention any developments for smart home products, such as smart speakers or smart displays, at this time. However, it is possible that the LLM smarts could eventually be extended to these devices as well.

Unfortunately, the internal restructuring has led to some team members being let go. Google has provided a 60-day period for those affected to find alternate positions within the company.

In a rapidly evolving landscape where technologies like ChatGPT and Bing Chat are gaining popularity, this leaked memo confirms that Google Assistant still has a future. By incorporating LLM technology, Google aims to make Assistant more powerful and capable of meeting peoples growing expectations for assistive and conversational technology.

View original post here:

Google Working to Supercharge Google Assistant with LLM Smarts - Fagen wasanni

SBU/Department:Hertfordshire Law School

FTE: 1 FTE working 37 hours per week Duration of Contract:Permanent Salary:AM1 64,946 - 71,305 per annum depending on skills and experience Location: De Havilland Campus, University of Hertfordshire, Hatfield

At Hertfordshire Law School we pride ourselves on delivering a truly innovative learning and teaching experience coupled with practice-led, hands-on experience. Our students consistently provide excellent feedback about their educational experience which is also evidenced through the number of students graduating with good honours degrees and our strong employability rates.

The School teaches Law (LLB and LLM) and Criminology (BA) programmes in a 10m purpose-built building on the University of Hertfordshire's de Havilland campus, which includes a full-scale replica Crown Court Room and state-of-the-art teaching facilities.

We are looking for an outstanding individual to provide academic leadership of the LLM Bar Practice Programme.

Main duties & responsibilities

The successful candidate will, in liaison with the Senior Leadership Team, manage and deliver the LLM Bar Practice Programme; monitor academic standards of the programme and ensure ongoing compliance with Bar Standards Board requirements. You will undertake the day-to-day management of the programme, including, as appropriate, the supervision of module leaders, identification of staffing needs, maintenance of programme documentation and records and provision of pastoral care.

Working closely with the Head of Department and Associate Deans, you will ensure the continuous development of the curriculum and act as chair of Programme Committees and relevant Examination Boards. You will support the marketing and recruitment of students and staff to the programme, both domestically and internationally, via the preparation of marketing and recruitment materials, organising and attending open days, international recruitment fairs and visiting collaborative partner institutions.

In addition, you will contribute to the delivery of the Schools co-curricular programmes and maintain and develop relationships with a wide range of Barrister Chambers and employers in the areas of legal and criminal justice practice to support the development of the programme and opportunities for students in Hertfordshire Law School.

Skills and experience needed

You will have proven experience as a programme leader or deputy programme leader of a professional law programme. Significant teaching experience of law on a Bar Professional Training Course/Programme in the UK within the last five years is essential. Ideally you will have experience as a practicing Solicitor or Barrister. You will also have demonstrable experience of programme/module design, with the ability to contribute to the design of engaging and intellectually stimulating modules and/or programmes. In addition, experience of line management of staff is desirable.

You will have an understanding of the Universitys strategic plan, regulations and processes and employability plans. You will be proficient in English, able to use technology to enhance delivery to students, have excellent organisation and self-management skills and the ability to negotiate with stakeholders. You will have a highly developed sense of professionalism and a commitment to student graduate success, including a commitment to equal opportunities and to ensuring that students from all backgrounds have the support they need to succeed and progress in their careers.

Qualifications required

You will have a good undergraduate degree or equivalent qualification, alongside a Master's qualification in law or equivalent professional qualification. A teaching qualification and / or Fellowship of AdvanceHE is desirable.

Additional benefits

The University offers a range of benefits including a pension scheme, professional development, family friendly policies, a fee waiver of 50% for all children of staff under the age of 25 at the start of the course, discounted memberships at the Hertfordshire Sports Village and generous annual leave.

How to apply

To find out more about this opportunity, please visit http://www.andersonquigley.com quoting reference AQ2099.

For a confidential discussion, please contact our advising consultants at Anderson Quigley: Imogen Wilde on +44 (0)7864 652 633, imogen.wilde@andersonquigley.com or Elliott Rae on +44 (0)7584 078 534, email elliott.rae@andersonquigley.com

Closing date: noon on Friday 1st September 2023.

Our vision is to transform lives and UH is committed to Equality, Diversity and Inclusion and building a diverse community. We welcome applications from suitably qualified and eligible candidates regardless of their protected characteristics. We are a Disability Confident Employer.

Original post:

Academic Manager / Programme Leader LLM Bar Practice job with ... - Times Higher Education