Privacy Policy


Get Certified Quickly

Home » News » Forum Systems Unveils Cutting-Edge GenAI Productivity-Risk Framework

Forum Systems Unveils Cutting-Edge GenAI Productivity-Risk Framework


Forum Systems, a leading provider of LLM and API technologies, has unveiled a significant advancement in the realm of Generative AI (GenAI) with the public release of two fine-tuned language models optimized for productivity-risk profiles. This announcement was made at the Gartner Data & Analytics Summit 2024 in London. The release includes Mistral QS-Sentry and Llama 3 QS-Sentry, both of which have undergone extensive fine-tuning to enhance their performance and align with enterprise needs.

The development of these fine-tuned models stems from Forum Systems’ dedication to addressing the critical challenge of balancing productivity and risk in GenAI deployments. Mamoon Yunus, CEO of Forum Systems, emphasized the importance of measuring the productivity-risk balance of AI models, noting that more restrictive models may enhance safety but could potentially hamper productivity. He highlighted the use of classic machine learning metrics such as precision and recall as proxies for productivity and risk evaluation.

The Mistral QS-Sentry and Llama 3 QS-Sentry models are based on Mistral-7B-Instruct-v0.2 and Meta-Llama-3-8B-Instruct, respectively. Forum Systems conducted thorough analyses to assess the productivity-risk profiles of these models before and after fine-tuning. The results revealed significant improvements in both models’ performance and alignment with enterprise objectives.

In addition to releasing the fine-tuned models, Forum Systems has also published two articles detailing its research findings and methodologies. The first article introduces a framework for balancing risk and productivity in GenAI deployments, while the second article examines the impact of fine-tuning on model performance. These articles provide valuable insights into the trade-offs involved in GenAI implementation and offer practical guidance for business leaders.

Forum Systems has made the Mistral QS-Sentry and Llama 3 QS-Sentry models available on Hugging Face, a platform for sharing AI models and datasets. By contributing to the broader research community, Forum Systems aims to facilitate collaboration and innovation in the field of AI governance and alignment.

Arun Chandrasekaran, a Gartner analyst, highlighted the transformative potential of Generative AI and emphasized the importance of responsible AI deployment. He recommended incorporating human feedback and design thinking into GenAI applications to promote harmonious interaction between humans and machines.

Forum Systems’ work underscores the growing importance of responsible AI deployment in enterprise settings. By optimizing productivity-risk profiles through fine-tuning, the company enables organizations to harness the benefits of GenAI while mitigating potential risks. This approach aligns with Gartner’s best practices for scaling Generative AI across the enterprise.

About the Gartner Data & Analytics Summit 2024:

The Gartner Data & Analytics Summits provide a platform for industry experts to discuss emerging trends and best practices in data and analytics. These summits offer valuable insights for business and technology leaders seeking to leverage data-driven strategies for competitive advantage.

About Forum Systems:

Forum Systems is at the forefront of the Enterprise GenAI revolution, offering innovative solutions for integrating GenAI technologies into corporate workflows. With its patent-pending QuantumSim™ technology and Forum Sentry platform, the company ensures the accuracy and reliability of GenAI output, enabling organizations to reimagine their business processes with AI.


Leave a Comment