Resource Library

Get a quick overview of the Fiddler AI Observability platform, including monitoring, explainability, fairness, and analytics.

The Fiddler Trust Service detects and monitors hallucinations, toxicity, PII leakage, and other LLM metrics, ensuring accurate, safe, and secure applications.

The Fiddler AI Observability platform helps financial service institutions monitor, analyze , explain, and improve predictive and generative AI models and applications.

See Fiddler in action and discover how Fiddler Trust Models detect and help you safeguard your LLM application while driving business growth.

Discover how Fiddler AI Observability’s advanced image explainability helps you gain deep insights into the predictions and behaviors of computer vision models.

The Fiddler AI Observability platform helps insurance companies monitor, analyze, explain, and improve predictive and generative AI models and applications.

Fiddler partners with federal agencies to ensure the safety and performance of AI models, enabling responsible AI with comprehensive observability and monitoring.

Fiddler provides computer vision capabilities to aid government agencies to detect and analyze issues in image models, enhancing high-stakes decision-making.

Fiddler's AI Observability platform aids government agencies by ensuring performance, transparency, and human-in-the-loop decision-making in AI applications.

Explore a method for measuring distributional shifts in text data using language model-based embeddings, highlighting the effectiveness of LLMs in capturing semantic relationships for this purpose.

Practical lessons and experiences on developing AI applications and chatbots using retrieval-augmented generation (RAG) methods.

Watch this keynote from Hanlin Tang on the discussion of the science and engineering behind LLMs in enterprise applications, the choice between buying LLM services from external providers and building custom models, specialized domain-specific models, and more.

Watch this expert panel discuss generative AI in enterprise, focusing on model efficiency, personalization, RAG's role, and aligning AI with business needs.

Watch this AI expert panel discuss the advancements and challenges in LLMs from proof-of-concept to production, the complexities of LLMOps, and regulatory compliance.

Watch this workshop to learn the importance of evaluating LLMs using Fiddler Auditor, strategies to mitigate prompt injection attacks, and developments in LLMs.

In this session, Stanford Professor James Zou discusses how ChatGPT's behavior is changing over time. Discover the fascinating research findings and implications for AI language models.

In this fireside chat with CEOs Krishna Gade and Lukas Biewald, they discuss AI trends, leveraging LLMs, and the complexities of LLM deployment at scale.

Watch this session as Josh Rubin explores the question of whether LLMs can be explained, emphasizing that understanding a model's behavior is crucial for identifying weaknesses, ensuring robustness, and aiding human decision-making.

In this workshop, Murtuza Shergadwala shares his experiences in building the RAG-based Fiddler Chatbot, and provides tips and tricks on prompt engineering, and more.

Watch this panel of industry leaders share their insights on LLM implementation and their significant impact on business ROI in their respective fields.

Watch this demo to explore how the Fiddler AI Observability Platform for MLOps enables data scientists and AI practitioners to monitor, explain, analyze, and improve ML models.

Watch this demo to explore how the Fiddler AI Observability Platform for LLMOps streamlines end-to-end LLMOps by enabling data scientists and AI practitioners.

In this workshop, see how Fiddler users can use the Fiddler Report Generator (FRoG) to create customizable reports for model risk management (MRM) and periodic model health reviews.

Improve LLMs with robustness testing before deploying them to production.

Learn how Fiddler’s unique, clustering-based method accurately monitors data drift in NLP models and LLM-based embeddings.

Learn more about Fiddler Auditor, the open source robustness library that facilitates red teaming of LLMs. Improve LLMs with robustness testing before deploying them to production.

Learn about Fiddler Auditor, the open source robustness library that facilitates red teaming of LLMs. Robustness testing is a critical step in pre-production to minimize hallucinations, bias, and adversarial attacks.

Get a quick overview of the critical role of AI Observability, the foundation you need to standardize LLM and MLOps practices, for building responsible AI.

Learn how to track the performance of LLM-based embeddings from OpenAI, Cohere, Anthropic, and other LLMs by monitoring drift using Fiddler.

Learn how Fiddler AI Observability provides ML teams with a unified platform to deliver and improve high-performing ML models while saving time and money.

Watch this panel session with Ricardo Baeza-Yates, Director of Research at Institute for Experiential AI, Northeastern University; Miriam Vogel, President and CEO at EqualAI, Chair at National AI Advisory Committee; Toni Morgan, Responsible Innovation Manager at TikTok, to learn why responsible AI principles and frameworks are necessary, why model governance is important, and the limitations of LLMs.

Saad Ansari, Director of AI at Jasper AI, dives into how AI can be used for societal benefit, the importance of intentionality when designing new technology, and applications for generative AI.

Hear from Amit Prakash, CTO and Co-founder at ThoughtSpot; Diego Oppenheimer, Partner and CEO in Residence at Factory; and Roie Schwaber-Cohen, Staff Developer Advocate at Pinecone on how LLMOps optimizes MLOps for LLMs, the key pieces of a generative AI workflow, and how ML teams can leverage LLMs in their applications.

Watch this session from Ali Arsanjani, PhD, Head of the AI Center of Excellence at Google, to learn about the importance of explainability, adaptability, and risk-minimizing in enterprise generative AI.

Watch this panel session with Payel Das, Principal Research Staff Member and Manager, Trusted AI a IBM Research; Srinath Sridhar, CEO and Co-founder at Regie.ai; Casey Corvino, CTO and Co-founder at Lavender AI to learn how engineering teams are using generative AI to improve productivity and how IBM is using generative AI to drive medical advances.

Watch this session with George Mathew, Managing Director at Insight Partners to learn about the current state of AI, the different modalities of human and machine interaction, the advancements, and challenges AI presents.

Learn how Fiddler AI Observability provides ML teams with a unified platform to monitor, analyze, explain, and improve ML models at scale, and build trust into AI.

Learn how Fiddler’s custom dashboards help ML teams obtain actionable model insights, and increase organizational alignment and collaboration.

Learn how to monitor models with unstructured data using Fiddler's cluster-based binning approach.

Learn how to perform root cause analysis to gain contextual insights, and improve model outcomes using Fiddler's analytics capabilities.

Learn how to evaluate and validate models in Fiddler before deploying them into production.

Learn how to get a deep understanding of model predictions using Fiddler's explainability capabilities.

Learn how to quickly detect model performance and drift issues, and reduce the time to troubleshoot issues with root cause analysis using Fiddler.

Learn how Fiddler’s powerful and customizable alerts provide early warning on model performance.

Krishnaram Kenthapadi, Chief Scientist at Fiddler AI, explains how model performance management enables real-time model monitoring, alerts, and explainability to help ML teams monitor, explain, analyze, and improve models, while operationalizing ML in a safe and trustworthy way.

Krishnaram Kenthapadi, Chief Scientist at Fiddler AI, explains how to monitor ML models in practice, featuring real-world use cases.

Hima Lakkaraju, assistant professor at Harvard University, summarizes the conclusions and key takeaways for ML model monitoring in practice.

Hima Lakkaraju, assistant professor at Harvard University, explains how to monitor the behavior of ML models through model interpretations and explanations.

Hima Lakkaraju, assistant professor at Harvard University, explains how to monitor the adversarial robustness of deployed ML models.

Krishnaram Kenthapadi, Chief Scientist at Fiddler AI, explains the common model issues that lead to performance degradation and how to monitor ML models.

Learn the human-centric challenges and requirements for ML model monitoring in real-world applications.

Learn about FairCanary, a novel approach to generating feature-level bias explanations that is an order of magnitude faster than previous methods.

A panel of industry and research experts discuss the state of explainable AI today and what the future holds.

Scott Zoldi, Chief Analytics Officer at FICO, has authored over 100 patents in ML and AI. He discusses why AI needs to grow up fast and what orgs can do about it.

Anjana Susarla, who holds the Omura-Saxena Professorship in Responsible AI at Michigan State, discusses the different dimensions of responsible AI.

Anand Rao, Global AI Lead, PwC discusses how organizations can implement AI responsibly.

Fiona McEvoy, founder of YouTheData.com, shares her perspective on algorithmic bias, deep fakes, emotional AI, and the way that AI systems impact our behavior.

Lofred Madzou, AI Lead at the World Economic Forum, shares his experiences managing AI projects around the world and the top things to keep in mind when implementing AI.

Karen Hao, Senior AI Reporter, MIT Technology Review discusses the current state of AI and democratizing its use.

Scott Belsky, Chief Product Officer, Adobe Creative Cloud discusses AI’s role in the creative world and its potential for unleashing the creative mind.

Listen to this episode of the Responsible AI podcast to hear how different companies are defining responsible AI and measuring the impact of AI applications.

Panelists discuss the increasing use cases for AI within finance, the unique regulatory and compliance considerations, and areas of opportunity.

Maria Axente, Responsible AI lead for PwC UK, shares what RAI means, why its importance is overlooked, and ways to incentivize teams to implement AI ethically.

Read about GroupShapley and GroupIG (Integrated Gradients), axiomatically justified methods to understand drift in ML model prediction distributions over time.

Improve your MLOps workflow on SageMaker using Fiddler's Model Performance Management platform.

Learn why fairness metrics need to incorporate intersectionality, including a simple method to expand the definitions of existing group fairness.

Watch this panel discussion on building responsible, ethical, and accountable AI.

Learn how organizations view and use explainability for stakeholder consumption, including a framework for establishing clear goals for explainability.

Learn how subtle differences in the underlying game formulations of existing methods can cause large differences in the attributions for a prediction.

Learn how randomized feature ablation helps measure the importance of a feature on a model's ability to make good predictions.

An overview of model interpretability and explainability in AI, key regulations/laws, and techniques/tools for providing explainability as part of AI/ML systems