Excited to see Dmitry joining this great lineup to share insights on real-world multimodal workflows. See you at the event!
Are you working with multimodal data? You won't want to miss this. These are three of the highest scoring submissions in our Track "𝗠𝘂𝗹𝘁𝗶𝗺𝗼𝗱𝗮𝗹 𝗦𝘆𝘀𝘁𝗲𝗺𝘀 𝗶𝗻 𝗣𝗿𝗼𝗱𝘂𝗰𝘁𝗶𝗼𝗻" Learning lessons we'll cover; 𝗖𝗹𝗼𝘀𝗲𝗱-𝗯𝗼𝘅 𝘃𝘀 𝗢𝗽𝗲𝗻-𝗯𝗼𝘅 𝗪𝗼𝗿𝗸𝗳𝗹𝗼𝘄𝘀 → Finer control, faster debugging, better calibration 𝗗𝗲𝗳𝗶𝗻𝗶𝗻𝗴 𝘁𝗵𝗲 𝗥𝗶𝗴𝗵𝘁 𝗘𝘃𝗮𝗹𝘀 → Human checkpoints for introspection and review 𝗗𝗮𝘁𝗮 𝗣𝗶𝗽𝗲𝗹𝗶𝗻𝗲 𝗕𝘂𝗶𝗹𝗱𝗶𝗻𝗴 𝗕𝗹𝗼𝗰𝗸𝘀 → Declarative pipelines, computed columns, batch execution 𝗪𝗵𝘆 𝗧𝗿𝗮𝗱𝗶𝘁𝗶𝗼𝗻𝗮𝗹 𝗔𝗽𝗽𝗿𝗼𝗮𝗰𝗵𝗲𝘀 𝗙𝗮𝗶𝗹 → Limits of text/image AI, need for agentic video methods 𝗩𝗶𝗱𝗲𝗼 𝗔𝗴𝗲𝗻𝘁 𝗔𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘂𝗿𝗲𝘀 → Planner–worker–reflector patterns for complex workflows 𝗣𝗿𝗮𝗰𝘁𝗶𝗰𝗮𝗹 𝗜𝗺𝗽𝗹𝗲𝗺𝗲𝗻𝘁𝗮𝘁𝗶𝗼𝗻 𝗦𝘁𝗿𝗮𝘁𝗲𝗴𝗶𝗲𝘀 → Transparent reasoning, multimodal handling, model orchestration 𝗕𝘂𝘀𝗶𝗻𝗲𝘀𝘀 𝗜𝗺𝗽𝗮𝗰𝘁 𝗮𝗻𝗱 𝗥𝗢𝗜 → Efficiency gains, identifying high-value use cases 𝗣𝗮𝗿𝗮𝗱𝗶𝗴𝗺 𝗦𝗵𝗶𝗳𝘁 𝗶𝗻 𝗠𝗲𝗱𝗶𝗮 𝗣𝗿𝗼𝗰𝗲𝘀𝘀𝗶𝗻𝗴 → Slice audio/video/sensor files into searchable segments for speed, cost, and accuracy I'm very proud to be hosting these incredible speakers. Denise Kutnick, CEO of Variata - over a decade of experience building multimodal systems and eval pipelines used by millions, from Intel Corporation to OctoAI (Acquired by NVIDIA) James Le - Head of Dev Experience TwelveLabs where they build multimodal foundation models., formerly Superb AI Inc. and The Full Stack Dmitry Petrov - entrpreneur + co-founder of DataChain, Inc. PhD in CompSci, creator of DVC -data version control (iterative.ai) We'll have food, drink, workshops, case studies and some other fun surprises planned. 📍 mlopsworld com Big thank you to our committee members for this track, and the chair Marcel Kornacker (Pixeltable) for your hard work here. Prakash Murugesan Ezequiel Lanza, MSc. Nima Eshraghi Vishakha Gupta Akshay Jain Dippu Singh Shubham Barhate Big thanks to some of our new community partners as well - please check them out! PyTexas Foundation Austin AI Alliance MLOps Club Marvelous MLOps Ask AI HACKAI