AIQCon

Check out our Co-Founder/CEO, Mohamed Elgendy's opening remarks at our inaugural AI Quality Conference (AIQCon). Mohamed discusses how to increase the adoption of AI products by developing innovative AI Quality processes. https://lnkd.in/ge2TMV8W

Missed the awesome AIQCon by MLOps Community & Kolena? Here are my notes: ??Themes ? ? Data quality and enabling data quality through tooling and thoughtful platform design ? The relationship of data quality to the success of generative AI projects ? The real, open-problems left to be solved with regards to classical ML and computer vision ? The challenges of getting more data: either through human labeling and annotation and synthetic data. ???Favorite Talks ?? ???12pm - EIGHTY-THOUSAND POUND ROBOTS: AI DEVELOPMENT & DEPLOYMENT AT KODIAK SPEED ?? I really enjoyed this case study presented by Kodiak’s Collin Otis — it can be rare to: ? Get a transparent look at the data → performance measurement tree, especially of autonomous vehicle companies; ? Not only get KPIs but also see what kind of automation and CI/CD practices are being implemented to support the data quality and automation processes; ? Wrapped up in a nicely packaged presentation with self-explanatory slides. ????12:50pm - OVERCOMING BIAS IN COMPUTER VISION AND VOICE RECOGNITION (w/ Skip Everling, Rajpreet Thethy, Doug Aley, Peter Kant) This is the talk I ended up taking the most notes on, especially because of the discussion around the role of high-quality labeling and annotations; the frustration of working with data labeling companies that are missing either services, best-of-breed tooling, or pricing; and the real scenarios where data quality (specifically a lack of the right data) could result in life or death scenarios (not to mention worsening wealth inequality in the US). ???1:55pm - GENERATING THE INVISIBLE: CAPTURING AND GENERATING EDGE-CASES IN AUTONOMOUS DRIVING (w/ Felix Heide) ? This was a really well-done deep-dive on how to use simulations to tackle the long-tail data challenges of autonomous vehicles: when you have a small sample of actual accidents, what needs to happen in order to train models to respond in critical scenarios? Coming out of CVPR and a number of conferences and events where synthetic data has started coming up again as an effective tool where counterfactual data doesn’t exist and labeling isn’t an option. I also enjoyed the thoughtful question by Jeremy Welland, Ph.D. on the use of near miss data as well as the insight by Feliz on the use of traffic and CHP reports. ?????2:55pm - PANEL: DATA QUALITY = QUALITY AI (w/ Sam Partee, Chad Sanderson, Joe Reis ??, Maria Zhang, Pushkar Garg) And of course, this panel of OG’s (as well as friends) on data quality and why we keep coming back to data quality, regardless of the level of “AI” the industry is hyping on. ?????Data, ML, and MLOps Fam One of the challenges of great events is sometimes you need to choose between attending a talk or catching up with the fam — it was exciting to catch-up (& meet IRL) with folks like: Wen Yang, Sadie St. Lawrence, Fanny Chow, Chandana Srinivasa, Christos Magganas, Stefan Krawczyk, Mihail Eric, David Scharbach, Faraz Thambi, Delia Lazarescu

Attending AIQCon, the AI Quality Conference — super excited, and we already have spotted Demetrios Brinkmann’s ukelele!

AIQCon is finally here! Watch Inductor's founder, Ariel Kleiner, live today @ 3:15 pm PDT as he discusses developer-first workflows for prototyping, evaluating, improving, and monitoring LLM apps. https://lnkd.in/gtEaMTjR Thank you MLOps Community for this opportunity to help developers build production-ready LLM apps!

Excited to listen to Stephanie Kirmer talk about balancing data privacy and innovation in AI/ML at AIQCon DataGrail

What's an insightful story you have about #AI quality? I'm going to share a few real-world anecdotes and lessons learned at AIQCon on Tue 6/25. If you’re working on #ML systems, you don’t want to miss this one! Use this code to get a 15% discount (”Community”) at https://lnkd.in/gk_J5uqq

In SF next week? Join us at AIQCon! Check out Jeremy Silva's talk -- details below and link in the comments. ??

Learn more about entity resolved knowledge graphs from Paco Nathan at AIQCon – the world's first AI Quality Conference – in SF on June 25th! Register with our 50% off code: SENZING https://lnkd.in/eD4EgXgB MLOps Community Kolena #EntityResolvedKnowledgeGraphs #EntityResolvedKnowledgeGraph #ERKG #EntityResolution #AIQCon #MLOps #AI #ArtificialIntelligence #MachineLearning

Join us at AIQCon for a thought-provoking discussion on one of the most pressing issues in artificial intelligence today: overcoming bias in voice and vision AI models. This panel moderated by Skip Everling (Head of DevRel of Kolena) will delve into the complexities of detecting, measuring, and mitigating bias to ensure fairness and inclusivity in AI systems. Rajpreet Thethy, Staff TPM at AssemblyAI, Doug Aley, CEO of Paravision, and Peter Kant, CEO of Enabled Intelligence, Inc will share their insights, best practices, and real-world examples, providing a comprehensive look at how leading companies are addressing bias in their AI technologies. Key points to be discussed include: -Identifying and Understanding Bias: Explore the different types of biases commonly encountered in voice and vision AI models and their impact on performance and fairness. -Detection Methods: Learn about the cutting-edge methods and tools used to detect bias, with examples of their effectiveness in real-world scenarios. -Measuring Bias: Understand the metrics and benchmarks used to measure bias and how these ensure comprehensive and accurate assessments. -Mitigating and Reducing Bias: Discover strategies and techniques for mitigating bias, supported by successful case studies from industry leaders. -Inclusive Data Practices: Gain insights into inclusive data collection and labeling practices that help avoid reinforcing existing biases. -Ethical Considerations: Delve into the ethical frameworks guiding AI development and how companies balance ethical, technical, and business requirements. -User Impact and Feedback: Learn how the impact of bias on different user groups is assessed and addressed through robust feedback mechanisms. -Real-World Applications and Challenges: Hear about real-world applications and the challenges faced in mitigating bias, with lessons learned from these experiences. -Future Trends and Developments: Look ahead to emerging techniques and technologies promising to further reduce bias, and understand the long-term vision for fair and unbiased AI systems. Register now and use code "KolenaFriend40" to save 40% off (limited quantities available). https://lnkd.in/gxTzZSXR

Join us at AIQCon on June 25th! Discover the future of AI with a focus on building reliable systems for enterprises. Vectara’s CEO, Amr Awadallah, will be speaking on setting new standards for AI quality. Use DISCOUNT CODE “Community” for a 15% discount. Register now: https://gag.gl/nEgbXa #AI #RAG