New York AI, Machine Learning and Computer Vision Meetup

Nov 11, 2026
9:00 AM - 11:00 AM PST
Online. Register for the Zoom!

Agentic RAG: Beyond Retrieve-and-Generate

Retrieval-Augmented Generation (RAG) has become the default pattern for grounding large language models in external knowledge, but most implementations still follow a rigid retrieve-once, generate-once pipeline — one that struggles with multi-hop questions, ambiguous queries, and knowing when its own retrieved context is insufficient. This talk introduces Agentic RAG, where retrieval is treated as an action within an agent's reasoning loop rather than a fixed upstream step.

We'll examine query decomposition and routing for breaking complex questions into targeted sub-queries, self-reflective and corrective retrieval loops that let an agent judge and re-query its own results, and tool-orchestration patterns (via MCP) that let retrieval sit alongside other agent actions like database lookups and API calls. Using a live architecture — evolving a standard RAG chatbot into an agentic, MCP-connected system — we'll walk through what changes in design, and where these systems introduce new failure modes: grounding drift, latency and cost from repeated retrieval loops, and cases where a simpler RAG pipeline still wins.

Attendees will leave with a practical framework for deciding when the added complexity of agentic RAG is worth it, and a set of design patterns for building it correctly.

About the Speaker

Balaji Venkatasubramaniyar is a Technical Lead at Wisdom Infotech, leading a 15+ person engineering team delivering enterprise solutions. With 13+ years of experience in enterprise software and insurance technology, he specializes in agentic AI systems, RAG architectures, and vector databases.

GeoAI for the Physical World: Earth Observation, Foundation Models, and Urban Digital Twins

Earth observation provides a unique form of computer vision for understanding the physical world at city to continental scales. In this talk, I will show how satellite imagery, geospatial data, machine learning, and foundation-model representations can be combined to characterize urban environments and environmental conditions.

I will present UrbanScope Open, an open GeoAI digital-twin prototype integrating Earth observation with 3D buildings, vegetation, land-surface temperature, air quality, noise, population, and other urban data. I will also share lessons from my research using geospatial foundation-model embeddings for environmental prediction across Europe.

The talk will discuss how these approaches can contribute to increasingly multimodal AI systems capable of reasoning about real-world environments.

About the Speaker

Cesar Alvarez is a researcher at the University of Augsburg working at the intersection of GeoAI, Earth observation, remote sensing, and environmental intelligence. His research applies machine learning, computer vision, and geospatial foundation models to problems including urban environments, climate risk, air quality, and agriculture.

Can agents get curious?

Most AI agents are good at answering a question once we tell them exactly what to look for. The harder problem is building agents that can explore a complex dataset autonomously: generating hypotheses, deciding which analyses are worth running, allocating additional compute when evidence is ambiguous, and knowing when they have enough evidence to stop.

In this talk, I’ll show an architecture for autonomous research agents that combines structured knowledge, iterative tool use, and explicit evidence tracking to turn open-ended questions into a sequence of testable investigations. I’ll discuss practical lessons from building and evaluating these systems, including why more test-time compute does not automatically produce better research and how provenance and evaluation can make long-running agents more reliable.

I’ll close with a live example of an agent exploring a dataset, revising its hypotheses, and choosing what to investigate next.

About the Speaker

Srivatsa P is a member of technical staff at Sigma Computing, where he works on AI agents that reason over complex enterprise data. Previously, he worked on machine learning at Apple and conducted research at the MIT Media Lab; outside of traditional ML, he has also worked on mapping coral reefs through underwater imaging, which sparked an enduring interest in how intelligent systems make sense of messy real-world data.

Does Cross-Validation Work in Deep Learning?

Cross-validation is a standard in hyperparameter optimization in classical ML, but is generally considered computationally prohibitive and unnecessary in deep learning. We investigated whether CV produces 1. better DL models 2. better estimates of future performance of said models. We focused on medical image classification as an important deep learning application domain.

Our results show that cross-validation consistently and materially reduces the gap between development and independent testing error for relatively small training datasets (up to about 1000 images). The test performance improvement was generally modest. We discuss the implications for deep-learning practitioners, including whether the increased computational cost of cross-validation is justified.

About the Speaker

Ljubomir Buturovic is an AI scientist and biotech executive with extensive experience in clinical diagnostics, including work at Inflammatix and Pathwork Diagnostics. His current research focuses on deep-learning image classification, infectious disease adjudication using LLMs, and development of AI models for helping critically ill patients.

Register to reserve your spot. Seats are limited.

Date, Time and Location

Nov 13, 2026
5:30 PM - 8:30 PM EST
Fractal, 111 Conselyea St, Fl 2, Brooklyn, NY 11211

Closing the human to robot gap for dexterous hands

Collecting task-specific robot data for multi-fingered hands is challenging due to the many difficulties that arise in teleoperation. That is why recently there has been a major focus on learning robot policies directly from human demonstrations. However, human demonstrations are difficult to work with; there is a major morphological and visual gap between human and robot hands, as well as between the environments they operate in.

In this talk, I'd like to discuss my efforts on closing this gap.

About the Speaker

Irmak Guzey is a rising 3rd year PhD student at New York University, currently advised by Lerrel Pinto. Her research focuses on robot learning for dexterous manipulation. She has been awarded a Fulbright scholarship and NYU's Best Master's Thesis Award.

Beyond Frames: Searching Hours of Footage with One Natural-Language Query

Most computer vision pipelines still treat video as a pile of frames, which makes 'find the moment where X happens' surprisingly hard at archive scale. This lightning talk shows how video-native embeddings and video-language models change that, turning hours of footage into something you can search with a single natural-language query. I'll cover how video is segmented and embedded, how retrieval works across visual, audio and text signals, where temporal context gets lost, and the evaluation problems teams hit when there's no clean ground truth. The examples come from field work deploying video understanding for media, news, streaming and sports organizations. You'll leave with a mental model and a checklist for building video search that holds up beyond the demo.

About the Speaker

Yash Gugale is a Senior Forward Deployed Engineer at TwelveLabs in New York, integrating video foundation models (Marengo for embeddings and search, Pegasus for video-language understanding) for media, news, streaming and sports customers. He previously spent about six years in video streaming engineering at Peloton and Samsung TV Plus.

Learning the Geometry of Compression: OrBIT

Most embedding compression methods impose a predefined coding geometry, such as coordinate partitions, low-rank subspaces, or unrestricted codebooks. In this talk, I introduce OrBIT, a framework that instead learns reusable geometric structure from operator dynamics to guide the compression of large language model embeddings. Using orbit-generated dictionaries, shared additive codewords, and redundant overlapping representations, OrBIT combines local compression with global reconstruction, allowing errors across representations to compensate for one another. I will explain the mathematical principles behind this construction, its compact decoder, and experiments demonstrating up to 37.9× compression on GPT-2 embeddings and over 23× on several 7B-model embedding tables, with competitive rate–distortion performance. The talk will explore how learning representation geometry can open new directions for efficient and structure-aware machine learning.

About the Speaker

Yunied Puig is a mathematician and machine learning researcher with a PhD specializing in functional analysis, operator theory, and linear dynamics. His research has appeared in leading mathematics journals, including the Journal of Functional Analysis, Mathematische Annalen, and Ergodic Theory and Dynamical Systems.

Building Real-World Computer Vision Systems with Voxel51

This talk will explore practical workflows for building, evaluating, and improving modern computer vision systems. We'll dive into real-world approaches to dataset curation, model analysis, multimodal AI workflows, and production-ready vision pipelines using open-source technologies.

The session is designed for engineers, researchers, and AI practitioners looking to better understand how teams are developing and scaling computer vision applications today. Expect practical demos, technical insights, and discussions around the evolving AI tooling ecosystem.

About the Speaker

Dan Gural leads technical partnerships at Voxel51, where he's building the Physical AI Workbench, a platform that connects real-world sensor data with realistic simulation to help engineers better understand, validate, and improve their perception systems.

Seats are limited, Pre-registration is mandatory.

Date, Time and Location

Nov 14, 2026
1:30 PM - 4:30 PM
NYU Kimmel Center - Room 914
60 Washington Square South
New York, NY 10012

Workshop Agenda
This hands-on session uses DROID, a real-world robotics dataset loaded into FiftyOne as a native multimodal MCAP recording, and YOLO11n, fine-tuned live during the session.

• Welcome + framing: from raw robot logs to a trained detector
• Explore a real DROID robotics recording in FiftyOne's native multimodal MCAP viewer: camera, proprioception, and language on one synced timeline, no ROS install required
• Curate: extract and browse frames from the recording, filter and deduplicate
• Compute embeddings on the curated frames; explore via similarity search and embeddings visualization (via Nebius Serverless AI Jobs)
• Auto-label: open-vocabulary detection to generate bounding boxes for the robot gripper and target objects
• Train: fine-tune a YOLO11n detector on the auto-labeled frames (via Nebius Serverless AI Jobs)
• Evaluate results and close the loop: view predictions back on the original MCAP timeline
• What else Nebius offers: Token Factory walkthrough — chat/vision models, fine-tuning, credits

Nov 19, 2026
9:00 AM - 10:00 AM PST
Online. Register for the Zoom!

Robot learning data doesn't look like the data that trains most vision-language models. It's temporal, low-dimensional in the parts that matter most, embodiment-specific, and expensive to collect, and increasingly, it's shipped in Hugging Face's LeRobot format. But packaging robot data consistently isn't the same as being able to see it: episodes hide broken metadata, silently dropped sensor modalities, and mismatched schemas between robots that look identical on paper.

Starting from the core differences that make robot data hard (temporal structure, proprioception, action grounding, embodiment heterogeneity, data scarcity), we'll move into live debugging: a metadata bug that broke multiple dataset imports, sensor streams that vanish without an error, and two robots running an identical benchmark whose data is structurally incompatible. By the end, you'll know what to check before you trust a LeRobot dataset, and how to actually look inside one.