BEGIN:VCALENDAR VERSION:2.0 PRODID:-//events//Events//EN CALSCALE:GREGORIAN X-WR-CALNAME:Nov 11 - AI\, ML and Computer Vision Meetup REFRESH-INTERVAL;VALUE=DURATION:PT1H X-PUBLISHED-TTL:PT1H BEGIN:VEVENT UID:event-xijs85du@ontown.app DTSTAMP:20261008T195827Z DTSTART:20261111T170000Z SUMMARY:Nov 11 - AI\, ML and Computer Vision Meetup LOCATION:San Francisco\, United States DESCRIPTION:AI Professionals San Francisco Machine Learning & Agentic AI\nJ oin our virtual meetup to hear talks from experts on cutting-edge topics a cross AI\, ML\, and computer vision.\n\nTime\, Date and Location\n\nNov 11 \, 2026\n9:00 AM - 11:00 AM PST\nOnline. Register for the Zoom! (https://v oxel51.com/events/ai-ml-and-computer-vision-meetup-november-11-2026)\n\nAg entic RAG: Beyond Retrieve-and-Generate\n\nRetrieval-Augmented Generation (RAG) has become the default pattern for grounding large language models i n external knowledge\, but most implementations still follow a rigid retri eve-once\, generate-once pipeline — one that struggles with multi-hop qu estions\, ambiguous queries\, and knowing when its own retrieved context i s insufficient. This talk introduces Agentic RAG\, where retrieval is trea ted as an action within an agent's reasoning loop rather than a fixed upst ream step.\n\nWe'll examine query decomposition and routing for breaking c omplex questions into targeted sub-queries\, self-reflective and correctiv e retrieval loops that let an agent judge and re-query its own results\, a nd tool-orchestration patterns (via MCP) that let retrieval sit alongside other agent actions like database lookups and API calls. Using a live arch itecture — evolving a standard RAG chatbot into an agentic\, MCP-connect ed system — we'll walk through what changes in design\, and where these systems introduce new failure modes: grounding drift\, latency and cost fr om repeated retrieval loops\, and cases where a simpler RAG pipeline still wins.\n\nAttendees will leave with a practical framework for deciding whe n the added complexity of agentic RAG is worth it\, and a set of design pa tterns for building it correctly.\n\nAbout the Speaker\n\nBalaji Venkatasu bramaniyar (https://www.linkedin.com/in/balaji-venkatasubramaniyar/) is a Technical Lead at Wisdom Infotech\, leading a 15+ person engineering team delivering enterprise solutions. With 13+ years of experience in enterpris e software and insurance technology\, he specializes in agentic AI systems \, RAG architectures\, and vector databases.\n\nGeoAI for the Physical Wor ld: Earth Observation\, Foundation Models\, and Urban Digital Twins\n\nEar th observation provides a unique form of computer vision for understanding the physical world at city to continental scales. In this talk\, I will s how how satellite imagery\, geospatial data\, machine learning\, and found ation-model representations can be combined to characterize urban environm ents and environmental conditions.\n\nI will present UrbanScope Open\, an open GeoAI digital-twin prototype integrating Earth observation with 3D bu ildings\, vegetation\, land-surface temperature\, air quality\, noise\, po pulation\, and other urban data. I will also share lessons from my researc h using geospatial foundation-model embeddings for environmental predictio n across Europe.\n\nThe talk will discuss how these approaches can contrib ute to increasingly multimodal AI systems capable of reasoning about real- world environments.\n\nAbout the Speaker\n\nCesar Alvarez (https://voxel51 .com/events/cesar.alvarez@uni-a.de) is a researcher at the University of A ugsburg working at the intersection of GeoAI\, Earth observation\, remote sensing\, and environmental intelligence. His research applies machine lea rning\, computer vision\, and geospatial foundation models to problems inc luding urban environments\, climate risk\, air quality\, and agriculture.\ n\nCan agents get curious?\n\nMost AI agents are good at answering a quest ion once we tell them exactly what to look for. The harder problem is buil ding agents that can explore a complex dataset autonomously: generating hy potheses\, deciding which analyses are worth running\, allocating addition al compute when evidence is ambiguous\, and knowing when they have enough evidence to stop.\n\nIn this talk\, I’ll show an architecture for autono mous research agents that combines structured knowledge\, iterative tool u se\, and explicit evidence tracking to turn open-ended questions into a se quence of testable investigations. I’ll discuss practical lessons from b uilding and evaluating these systems\, including why more test-time comput e does not automatically produce better research and how provenance and ev aluation can make long-running agents more reliable.\n\nI’ll close with a live example of an agent exploring a dataset\, revising its hypotheses\, and choosing what to investigate next.\n\nAbout the Speaker\n\nSrivatsa P (https://www.linkedin.com/in/srivatsa-p) is a member of technical staff a t Sigma Computing\, where he works on AI agents that reason over complex e nterprise data. Previously\, he worked on machine learning at Apple and co nducted research at the MIT Media Lab\; outside of traditional ML\, he has also worked on mapping coral reefs through underwater imaging\, which spa rked an enduring interest in how intelligent systems make sense of messy r eal-world data. URL:https://ontown.app/e/xijs85du-nov-11-ai-ml-and-computer-vision-meetup/ END:VEVENT END:VCALENDAR