Skip to main content

Monthly Archives: August 2024

Powering Reliable and Efficient AI Systems

Posted on August 30, 2024 - 12:34 am by
aligned AI

The following blog post was originally published by Hitachi Ventures on Medium, highlighting their investment in Trustwise.  

In the rapidly evolving landscape of artificial intelligence (AI), ensuring the performance, safety, and ethical deployment of AI systems is paramount. We are excited to announce our participation as the lead investor in the seed financing of Trustwise, a pioneer in generative AI application performance and risk management, along with Allstate Strategic Ventures, Firestreak Ventures and Grit Ventures.

Founded by Manoj Saxena (former GM of IBM Watson and 3x entrepreneur), Trustwise’s innovative approach to AI safety, cost efficiency, and environmental sustainability aligns perfectly with Hitachi Ventures’ focus on supporting responsible AI innovation. We are confident that Trustwise has the ability to unlock the full potential of generative AI for enterprises across various sectors by being the foundation of trust. Optimize:ai by Trustwise is a pioneering solution for managing performance, risk and environmental impact of AI. It offers developers an API to mitigate risks and cut costs linked to deploying Large Language Models (LLMs) in critical enterprise settings.

Trustwise is Addressing Critical AI Challenges

aligned AI

1) Cost Efficiency

The operational costs associated with AI, particularly LLMs, can be prohibitive for a lot of enterprises. Trustwise Optimize:ai addresses this by routing the right prompt to the right LLM and reducing LLM operational costs by up to 80%. This significant cost reduction is achieved through advanced optimization techniques and dynamic model selection, making AI deployments more economically viable for enterprises.

2) Safety and Compliance

One of the most significant hurdles in deploying AI, particularly large language models (LLMs), is ensuring safety and compliance. Trustwise Optimize:ai excels in this area by providing robust AI safety and alignment controls. The platform actively mitigates risks such as hallucinations and data leakage, which are common issues with generative AI models. Optimize:ai not only adheres to internal corporate AI policies but also complies with major AI standards and regulations including EU AI Act, NIST AI RMF 1.0, RAISE Safety and Alignment Benchmarks by the Responsible AI Institute, and the SCI ISO software carbon intensity standard.

3) Environmental Sustainability

AI’s environmental impact is a growing concern. Training and large volumes of inference of models can result in substantial carbon emissions. Trustwise Optimize:ai leverages impact data from cloud providers and model providers to understand the impact of each model response and suggests the best alternative to reduce the carbon footprint of AI deployments by up to 64%.

What is Trustwise Optimize:ai and how does it work?

optimize AI

Trustwise employs rigorous red-teaming and optimization of generative AI workloads through advanced APIs, datasets, and AI safety controls. This approach protects enterprises from potential AI failures and excessive expenses while having a strong understanding of the carbon footprint of their AI deployments.

Powered by THEO (Trustworthy High-Efficiency Optimizer), an optimization algorithm that continuously monitors AI safety and alignment, performs red-teaming and dynamically selects models for large-scale generative AI operations, supporting diverse AI models and architectures and ensuring that AI systems are both high-performing and safe all through a single API that integrates seamlessly with any AI model and cloud environment. This flexibility eliminates the risk of vendor lock-in and ensures that enterprises can leverage the best AI technologies without high switching costs and compatibility issues. This universal adapter approach is crucial for enterprises adopting AI across diverse platforms, environments and use cases.

Your AI Use case

Increasing value creation for enterprises

Trustwise has already demonstrated a significant impact across various industries specially across financial and healthcare institutions.

For example:

  • Lloyds Banking Group reported that Trustwise Optimize:ai significantly reduced their generative AI operational costs and carbon footprint while ensuring compliance with internal policies and external regulations.
  • NHS highlighted the platform’s ability to enhance patient care through reliable and accurate AI solutions, reinforcing the importance of trust and safety in healthcare AI applications.
  • Lyric Health emphasized the role of Trustwise in driving AI adoption in healthcare, ensuring cost control and compliance risk minimization.

Trustwise is not just a technological solution; but can be a strong strategic partner in our journey towards safe, efficient, and sustainable AI adoption.

We look forward to the continued success and impact of Trustwise.

CEO Open Letter

Posted on August 14, 2024 - 8:30 pm by

Trustwise: Unlocking AI’s Potential with Trust

Saxena Manoj ORIGINAL 682x830
Manoj Saxena

Generative AI was supposed to revolutionize productivity, creativity, and innovation across industries.

The reality has not yet quite lived up to the hype.

Why? Because throughout history, it’s the trustworthiness of products, beyond mere technological innovation, that has truly shaped our world.

Torches harnessed fire, circuit breakers secured electrical appliances, and secure payment systems enabled internet commerce.

Similarly, Generative AI is on a journey toward adaptive safety and reliability. While it holds tremendous potential, it also presents challenging and potentially unsafe outcomes that must be addressed.

Trustwise is building the trust layer for generative AI making AI products reliable and safe. Without this layer, AI cannot achieve its full potential.

We believe a trusted AI system operates safely and responsibly, delivers accurate and beneficial outcomes, complies with regulations and company policies, and ensures cost and carbon efficiency.

For instance, a trusted healthcare AI diagnoses and recommends treatments without bias. It protects patient data and complies with regulations. A trusted banking AI advisor provides accurate financial advice, managing transactions securely. It adheres to banking policies and regulations and is cost and carbon efficient.

Our Optimize:ai API ensures AI alignment with corporate policies and industry regulations and safeguards AI outputs from sensitive data leakage and hallucinations. It seamlessly integrates with enterprise systems to reduce token consumption and carbon footprint by up to 80%. Unlike other competitors, Trustwise ensures AI systems are not just observable or governable but also safe, steerable, and both cost and carbon-efficient.

I’m excited for a future where trusted AI drives unprecedented productivity, creativity, and innovation. I believe Trustwise is leading the way to this future by making AI applications human-centric and trustworthy, empowering businesses to innovate confidently and efficiently with generative AI.

I’m incredibly proud of the Trustwise team. Over the last 18 months, we have doubled our revenues, tripled our number of active engagements and headcount, and laid the groundwork for customers to redefine how they build and innovate with trustworthy AI systems.

But I’m even more proud of the mission we wake up every day to tackle: At Trustwise, we believe we’re shaping the future of technology for the next 100 years. We’re not just building a software product; we’re creating a foundation of trust and innovation that will redefine how AI integrates into business and society to enhance our lives. We’re not settling for anything less.

If you share our vision and commitment to building human-centric, trustworthy AI, come join us in our journey.

Onwards!

Manoj Saxena

CEO and Founder

Trustwise Joins NVIDIA Inception Program

Posted on August 13, 2024 - 3:47 pm by
nvidia trustwise

At Trustwise, we build software that mitigates evolving AI risks and challenges so our customers can gain competitive advantages from AI while lowering their operational costs, increasing security and governance, and reducing carbon emissions. To further accelerate our innovation and growth, we’re proud to announce that we’ve been accepted into NVIDIA Inception, a program that nurtures startups revolutionizing industries with technological advancements. 

The Inception program will provide Trustwise with access to NVIDIA’s cutting-edge technology and go-to-market support to help deliver NVIDIA’s Inference Microservice (NIM) platform-based generative AI (GenAI) solution to customers, enabling them to rapidly develop their GenAI systems. The program also gives Trustwise the opportunity to collaborate with industry-leading experts and other AI-driven organizations.

Trustwise Optimize:ai accelerates the development of trustworthy GenAI systems while enabling enterprises to run applications and workloads efficiently across cloud, data center, and edge infrastructures. Optimize:ai, delivered through a single API, provides developers with four NIM microservices to reduce operational costs and carbon emissions:

  • Safe:ai NIM reduces risk of hallucinations and data leaks in GenAI systems by up to 20x using fine-tuned models for industry-leading hallucination detection.
  • Align:ai NIM ensures that GenAI systems are in compliance with corporate policies, industry, and governmental regulations, including the European Union AI Act, NIST AI RMF 1.0, Responsible AI Institute RAISE Safety and Alignment Benchmarks, and SCI ISO software carbon intensity standard.
  • Efficient:ai NIM optimizes the costs associated with developing GenAI systems by using advanced caching, chunking, and pareto optimization techniques and deliver up-to 5x reductions.
  • Green:ai NIM helps understand and reduce carbon impact of AI systems by up to 3x using SCER rating along with carbon maps for data centers globally.

Trustwise’s customers benefit from the company’s participation in Inception as it helps relieve their developers from the hassles of creating custom microservices or the complexity of third party integrations that might not be compatible with their infrastructure.

“As a participant in the NVIDIA Inception program, Trustwise is excited to leverage NIM to ensure our Optimize:ai solutions run efficiently across various enterprise environments, including cloud, data centers, and edge. NVIDIA NIM’s ability to handle the complexity of modern AI workloads aligns perfectly with our mission to deliver advanced AI applications that require robust, reliable, and efficient performance. Our participation in NVIDIA enables Trustwise to continue innovating with cutting-edge technology,” said Manoj Saxena, founder and CEO of Trustwise.

NVIDIA Inception helps startups during critical stages of product development, prototyping, and deployment. Every Inception member gets a custom set of ongoing benefits, such as NVIDIA Deep Learning Institute credits, preferred pricing on NVIDIA hardware and software, and technological assistance, which provides startups with the fundamental tools to help them grow.

Visit this page and sign up for our First Optimization for Free program and schedule a demo.