BEGIN:VCALENDAR VERSION:2.0 PRODID:-//CalDiscovery//caldiscovery//EN CALSCALE:GREGORIAN METHOD:PUBLISH DESCRIPTION:Discovered and maintained by CalDiscovery (https://caldiscover y.com). Event details belong to their publishers. Contains information fro m CalDiscovery\, which is made available under the ODbL 1.0. SOURCE:https://caldiscovery.com/santa-cruz-ca/src-15923.ics X-PUBLISHED-TTL:PT1H X-WR-CALDESC:Discovered and maintained by CalDiscovery (https://caldiscove ry.com). Event details belong to their publishers. Contains information fr om CalDiscovery\, which is made available under the ODbL 1.0. X-WR-CALNAME:CalDiscovery - santa-cruz-ca BEGIN:VEVENT SUMMARY:Sept 17 - ADAS\, AV\, and AI Meetup DTSTART:20260917T160000Z DTEND:20260917T180000Z DTSTAMP:20260819T003306Z UID:event_316174902@meetup.com CATEGORIES:tech DESCRIPTION:Santa Cruz AI and Machine Learning Meetup Group\nJoin our virt ual meetup to hear talks from experts on cutting-edge topics across AI\, M L\, and computer vision.\n\n**Time\, Date and Location**\n\nSep 17\, 2026\ n9:00 AM - 11:00 AM PST\nOnline. **[Register for the Zoom!](https://voxel5 1.com/events/adas-av-and-ai-meetup-september-17-2026)**\n\n**AI for Autono mous Driving: From Data to Decisions**\n\nBuilding reliable automated driv ing systems is as much a data and engineering challenge as a modeling one. In this talk\, Tin will share perspectives from his work at Porsche AG on applying modern AI methods across the autonomous driving development proc ess\, from making sense of large-scale driving data to understanding and e valuating how AI-based systems behave on the road. He'll discuss lessons l earned from real-world development\, where today's approaches shine\, and where hard problems remain for the ADAS and AV community.\n\n*About the Sp eaker*\n\n[Tin Stribor Sohn](https://www.linkedin.com/in/tin-stribor-sohn/ ) is a PhD Student at Porsche AG and Karlsruhe Institute of Technology in the area of Foundation Models for Scenario Understanding and Decision Maki ng in Autonomous Robotics\, Tech Lead at Data Driven Engineering for Auton omous Driving\, Prior: Master in CS at University of Tuebingen with focus on Computer Vision and Deep Learning and co-founder of a software company for smart EV charging\n\n**Advancing ADAS and Autonomous Vehicle Developme nt with Multimodal Data**\n\nADAS and autonomous vehicle systems rely on i ncreasingly complex data from cameras\, video\, LiDAR\, radar\, and other sensor streams. In this session\, Murilo will introduce Voxel51 and explor e how the latest multimodal capabilities in FiftyOne help teams bring thes e data sources together to better understand their datasets and model beha vior. He’ll discuss how unified workflows for visualization\, search\, c uration\, and evaluation can help ADAS and AV teams uncover challenging sc enarios\, investigate model failures\, and build safer\, more reliable aut onomous systems.\n\n*About the Speaker*\n\n[Murilo Gustineli](https://www. linkedin.com/in/murilo-gustineli/) is a Machine Learning Engineer at Voxel 51 working at the intersection of representation learning and computer vis ion. He holds an M.S. in Computer Science from Georgia Tech\, where he co- leads the DS@GT Applied Research & Competitions group\, advancing machine learning research through competitive challenges and peer-reviewed publica tions.\n\n**From Survey-Grade Maps to Physical AI: Scaling Real-World Data for Training and Simulation**\n\nPhysical AI systems are increasingly con strained not by model architectures\, but by the availability of scalable\ , high-fidelity real-world data. This talk explores how Dynamic Map Platfo rm transforms survey-grade road assets collected across 1.8 million km of roads worldwide into training- and simulation-ready datasets\, including p oint clouds\, imagery\, HD maps\, road surface models\, and 3D Gaussian Sp latting representations.\n\nWe will discuss why geometric accuracy\, seman tic understanding\, and real-world diversity are critical to building robu st autonomous driving systems. Attendees will learn how real-world geospat ial data can be structured and scaled for AI training and simulation workf lows.\n\n*About the Speaker*\n\n[Ryoto Miyake](https://www.linkedin.com/in /ryoto-miyake/) is a Software Engineer at Dynamic Map Platform\, where he works on transforming large-scale geospatial data into AI-ready datasets f or training\, simulation\, and validation\, such as HD maps and 3D Gaussia n Splatting. With a background in transportation engineering\, he works cl osely with automotive manufacturers and industry partners to bridge large- scale real-world mapping data with next-generation AI and mobility systems . LAST-MODIFIED:20260819T003306Z URL:https://www.meetup.com/santa-cruz-machine-learning-meetup-group/events /316174902/ X-SOURCE-NAME:Santa Cruz AI and Machine Learning Meetup Group END:VEVENT BEGIN:VEVENT SUMMARY:Sept 22 - FiftyOne Agent: Automate Visual AI Workflows with Natura l Language DTSTART:20260922T160000Z DTEND:20260922T180000Z DTSTAMP:20260815T030927Z UID:event_316064261@meetup.com CATEGORIES:tech DESCRIPTION:Santa Cruz AI and Machine Learning Meetup Group\nLearn about F iftyOne Agent\, an AI assistant built directly into FiftyOne that turns pl ain-language requests into real dataset operations.\n\n**Date\, Time and L ocation**\n\nSep 22\, 2026\n9:00 AM - 10:00 AM Pacific\nOnline. **[Registe r for the Zoom!](https://voxel51.com/events/meet-the-fiftyone-agent-automa te-visual-ai-workflows-with-natural-language-september-22-2026)**\n\nAsk i t to find and remove duplicate images\, run object detection and surface l ow-confidence predictions\, or evaluate a model and summarize where it fai ls\, and the agent handles execution end to end.\n\nWe will also walk thro ugh the newest capabilities shipping with this release\, including code ge neration and plugin generation. You will see how to go from a conversation al prompt to a custom dashboard\, visualization\, or full FiftyOne applica tion\, and how to package multi-step workflows as reusable skills the agen t can call on demand.\n\nBecause the agent runs inside FiftyOne's secure g uardrails\, teams keep full control. Connect your preferred models from ov er 100 LLM providers\, route requests through your own enterprise gateway\ , and maintain audit logging and user attribution for every action the age nt takes. LAST-MODIFIED:20260815T030927Z URL:https://www.meetup.com/santa-cruz-machine-learning-meetup-group/events /316064261/ X-SOURCE-NAME:Santa Cruz AI and Machine Learning Meetup Group END:VEVENT BEGIN:VEVENT SUMMARY:Sept 24 - AI\, ML and Computer Vision Meetup DTSTART:20260924T160000Z DTEND:20260924T180000Z DTSTAMP:20260815T030927Z UID:event_315388226@meetup.com CATEGORIES:tech DESCRIPTION:Santa Cruz AI and Machine Learning Meetup Group\nJoin our virt ual meetup on September 24 to hear talks from experts on cutting-edge topi cs across AI\, ML\, and computer vision.\n\n**Date\, Time and Location**\n \nSep 24\, 2026\n9:00 AM - 11:00 AM PST\nOnline. **[Register for the Zoom! ](https://voxel51.com/events/ai-ml-and-computer-vision-meetup-september-24 -2026)**\n\n**How Do Mercedes-Benz AI Principles Drive our Innovation?**\n \nAt Mercedes-Benz\, our AI Principles guide every step of innovation\, em phasizing responsible use\, safety and reliability\, explainability\, and the protection of privacy. These principles go beyond statements and activ ely shape how we design\, test\, and deploy AI systems in real-world autom otive and enterprise settings. In this talk\, I will present how these pri nciples inspired our recent research on when reusing LoRA (Low-Rank Adapta tion) is effective. By combining theoretical analysis with synthetic data as a proxy for enterprise scenarios\, we uncovered the strengths and limit ations of modular AI components under constrained data access. Our finding s provide practical guidance on when reused LoRAs could deliver high-quali ty results.\n\n*About the Speaker*\n\n[Mei-Yen Chen](https://www.linkedin. com/in/mei-yen-chen-22937787) is a Senior Data Scientist at Mercedes-Benz Tech Innovation GmbH in Germany with 10 years of industry experience in AI and data solutions. She leads early-stage AI projects across business fun ctions and collaborates with research institutions on machine learning and responsible AI.\n\n**Region Tokens as the Visual Primitive: From Recognit ion to World Modeling**\n\nPatch-based tokenization has become the default interface between vision encoders and downstream models\, yet patches car ry no semantic structure and scale poorly with resolution and temporal ext ent. This talk presents a research program centered on replacing patch tok ens with region-level representations — semantically dense tokens ground ed in visual entities rather than arbitrary grid crops.\n\nI will describe RELOCATE\, REN\, and T-REN\, a progression of methods that produce region tokens via pooling\, train them with region-level objectives\, and extend them to video with temporal coherence. I will then present ongoing work i ntegrating region tokens into VLMs to directly expand visual context capac ity\, and preliminary results on future region trajectory prediction as a foundation for world modeling.\n\nThe broader thesis is that region-level tokens are a more natural unit of visual computation than patches\, and th eir advantage compounds as task complexity\, resolution\, and temporal hor izon increase.\n\n*About the Speaker*\n\n[Savya Khosla](https://www.linked in.com/in/savyakhosla/) is a second-year Ph.D. student at the University o f Illinois Urbana-Champaign\, advised by Prof. Derek Hoiem and Prof. Alex Schwing.\n\n**Leveraging Text-To-Image Diffusion Models for Consistent Set -to-Set Generation**\n\nImage collections are humans' primary way of captu ring the world\, yet advances in generative editing remain largely inappli cable to this modality. We address this gap by introducing Match-and-Fuse - a zero-shot\, training-free method for consistent set-to-set generation from image collections that share a common visual element but differ in vi ewpoint\, capture time\, and surrounding content.\nOur key idea is a unifi ed graph-based framework that combines dense correspondences with an emerg ent prior in text-to-image diffusion models to generate coherent canvases. We achieve state-of-the-art consistency and visual quality\, and unlock n ew creative capabilities for content generation.\n\n*About the Speaker*\n\ n[Kate Feingold](https://www.linkedin.com/in/katefeingold/) is a PhD stude nt in Computer Vision at the Weizmann Institute of Science. Her research s its at the intersection of generative models\, 3D/4D perception\, and mult imodal learning\, focusing on problems where vision meets other modalities or paradigms in creative tasks.\n\n**Yield Estimation of a Coffee in a de nse environment**\n\nThis presentation provides a detailed workflow relate d to coffee yield estimation in a dense environment. With photos of pre-ha rvest coffee plants from a couple of coffee estates\, details related to p re-processing\, annotation to detect regions of interest (ROI)\, object de tection training and inferencing results with various Yolo models and fina lly segmentation with SAM2 and Yolo\\*-seg with training and inference res ults to determine the count of raw\, pre-mature\, mature and over-mature c offee berries and finally the yield of the entire estate. All this is base d on real world data captured on iPhone and android phones.\n\n*About the Speaker*\n\n[Raghu M. Rao](http://www.linkedin.com/in/raghumrao) is a cons ultant working on applications of computer vision AI models. He was previo usly with AMD and Xilinx. He has a Ph.D. in Wireless Communications from U CLA and is a Senior Member\, IEEE. His current interests are in applicatio ns of AI for agriculture\, health care and wireless communications. LAST-MODIFIED:20260815T030927Z URL:https://www.meetup.com/santa-cruz-machine-learning-meetup-group/events /315388226/ X-SOURCE-NAME:Santa Cruz AI and Machine Learning Meetup Group END:VEVENT BEGIN:VEVENT SUMMARY:Sept 30 - Building Composable Vision Workflows in FiftyOne DTSTART:20260930T160000Z DTEND:20260930T180000Z DTSTAMP:20260815T030927Z UID:event_316064668@meetup.com CATEGORIES:education,tech DESCRIPTION:Santa Cruz AI and Machine Learning Meetup Group\nThis workshop explores the [FiftyOne](https://docs.voxel51.com/index.html) plugin frame work to build custom computer vision applications. You’ll learn to exten d the FiftyOne App with Python based panels and server side operators\, as well as integrate external tools for labeling\, vector search\, and model inference into your dataset views.\n\n**Date\, Time and Location**\n\nSep 30\, 2026\n9 AM - 10 AM PST\nOnline. **[Register for the Zoom!](https://v oxel51.com/events/building-composable-vision-workflows-in-fiftyone-septemb er-30-2026)**\n\nYou’ll also automate repetitive tasks by writing custom workflows executing within the FiftyOne environment. Attendees will learn to transform FiftyOne from a visualization tool into a central hub for yo ur vision stack.\n\nWhat you'll learn:\n\n* **Build Python plugins.** Defi ne plugin manifests and directory structures to register custom functional ity within the FiftyOne ecosystem.\n* **Develop server side operators.** W rite functions to execute model inference\, data cleaning\, or metadata up dates from the App interface.\n* **Build interactive panels.** Create cust om UI dashboards using to visualize model metrics or specialized dataset d istributions.\n* **Manage operator execution contexts.** Pass data between the App front end and your backend to build dynamic user workflows.\n* ** Implement delegated execution.** Configure background workers to handle lo ng running data processing tasks without blocking the user interface.\n* * *Build labeling integrations.** Streamline the flow of data between FiftyO ne and annotation platforms through custom triggers and ingestion scripts. \n* **Extend vector database support.** Program custom connectors for exte rnal vector stores to enable semantic search across large sample datasets. \n* **Package and share plugins.** Distribute your extensions internally a nd externally LAST-MODIFIED:20260815T030927Z URL:https://www.meetup.com/santa-cruz-machine-learning-meetup-group/events /316064668/ X-SOURCE-NAME:Santa Cruz AI and Machine Learning Meetup Group END:VEVENT BEGIN:VEVENT SUMMARY:Oct 8 - MCP\, Agents and Skills Meetup Meetup DTSTART:20261008T160000Z DTEND:20261008T180000Z DTSTAMP:20260815T030927Z UID:event_316022265@meetup.com CATEGORIES:tech DESCRIPTION:Santa Cruz AI and Machine Learning Meetup Group\nJoin our virt ual meetup to hear talks from experts on MCP\, agents and skills.\n\n**Dat e\, Time and Location**\n\nOct 08\, 2026\n9:00 AM - 11:00 AM PST\nOnline. **[Register for the Zoom!](https://voxel51.com/events/mcp-agents-skills-me etup-october-8-2026)**\n\n**Designing Multi‑Agent Systems: Sequential\, Parallel\, and Beyond with ADK**\n\nMulti‑agent systems are powerful but choosing the wrong interaction pattern can quickly lead to fragile\, slow \, or expensive AI systems.\nIn this talk\, we explore the core multi‑ag ent design patterns enabled by ADK\, including sequential\, parallel\, and more advanced coordination models. Rather than focusing on tools alone\, we’ll look at how to think architecturally about agent collaboration.\n\ nYou’ll learn:\n\n* When sequential agents are the right choice and when they become a bottleneck\n* How parallel agents improve speed and coverag e (and the trade‑offs they introduce)\n* Common failure modes in poorly designed agent interactions\n* Practical criteria for choosing the right p attern based on task\, latency\, and reliability\n\nBy the end of the sess ion\, you’ll have a clear mental model for designing multi‑agent syste ms that are intentional\, scalable\, and production‑ready.\n\n*About the Speaker*\n\n[Dr Roushanak Rahmat](https://www.linkedin.com/in/roushanakra hmat/) is an Enterprise AI Architect\, Google Developer Expert (AI & Cloud )\, and recognized among the Top 100 Women in Tech (2025). With a PhD in A rtificial Intelligence and over 15 years of experience\, she specializes i n designing and delivering enterprise-scale Generative AI\, Agentic AI\, a nd Deep Learning solutions that transform industries including healthcare\ , finance\, energy\, and public services.\n\n**Privacy by Deployment: Arch itecting Agent-Driven Localization Workflows for Regulated Environments**\ n\nMost enterprise AI today is private by promise - a DPA\, a SOC 2 report \, or a contract clause that says\, "we won't train on your data". For a r egulated buyer\, these are remedies after a breach\, not controls that pre vent or contain one. For organizations in healthcare\, finance\, defense\, and government\, privacy often requires stronger guarantees: data residen cy\, customer-controlled execution\, and\, in some cases\, operation withi n air-gapped environments.\n\nThis session demonstrates how agentic AI can automate a localization workflow while operating within these constraints . Using a real-world localization pipeline as an example\, we will show ho w agentic systems can coordinate translation\, review\, quality assurance\ , and content preparation tasks while incorporating human checkpoints for approval and oversight.\n\nWe will also walk through the architectural pat terns that enable these workflows to run inside customer-controlled and ai r-gapped environments without transferring sensitive content outside the c ustomer boundary. The session includes a live product demonstration.\n\nKe y Takeaways\n\n* Architectural patterns for deploying agentic AI in air-ga pped and customer-controlled environments\n* How agentic systems can autom ate localization workflows while preserving critical human review and appr oval processes\n* Practical considerations for operating agentic workflows in regulated environments with auditability and governance requirements\n \n*About the Speaker*\n\n[Shruti Joshi](https://www.linkedin.com/in/28shru ti) is building an AI powered secure localization stack for regulated indu stries such as healthcare\, legal\, finance that cannot send their content to a typical hosted SaaS. She brings 12+ years of engineering and archite cture experience to the question this talk addresses: how do you make an a gentic AI system deployable inside a regulated perimeter.\n\n**MCP Is the Interface\; Skills Are the Operating Discipline**\n\nThis talk shows how M CP and Agent Skills work together in practical agent systems. MCP gives ag ents a standard interface to tools\, data\, and workflows\; skills encode the operating discipline that makes those connections reliable. Using a sa nitized field-operations ledger as the case study\, the talk walks through source intake\, normalized state\, uncertainty labels\, role prompts\, QA gates\, and share-safe status drafting.\n\n*About the Speaker*\n\n[Chuck Hernandez](https://www.linkedin.com/in/chuck-hernandez/) is an AI engineer ing and client-delivery leader with 10+ years across software\, data platf orms\, and enterprise implementation\, including 3+ years shipping product ion GenAI systems.\n\n**Agentic engineering is about good guidance.**\n\nG arbage Inn. Is garbage out? This is true. For many input and output proces ses. In biological life and in computer systems\, and equally true when wo rking with LLM’s. The better the prompt\, the better the context\, the b etter the focus\, And the better the contextual awareness\, the better the quality of the output the LLM’s generates.\nThis is the governance\, ar t and practice of what we like to call agentic engineering\, something I'v e been practicing over the last year.\n\n*About the Speaker*\n\n[Dimitri G eelen](https://www.linkedin.com/in/dimitrigeelen/) builds things that don' t need him once they're done. Frameworks\, transitions\, agentic systems — the measure of success is always the same: does it hold up when he lea ves the room? He understands not just how to deploy\, but what it takes fo r a new service to survive and scale inside a complex enterprise. LAST-MODIFIED:20260815T030927Z URL:https://www.meetup.com/santa-cruz-machine-learning-meetup-group/events /316022265/ X-SOURCE-NAME:Santa Cruz AI and Machine Learning Meetup Group END:VEVENT END:VCALENDAR