AI/ML

Sep 24, 2026 · 20:48

One Operator, Many Drones: Inside Skydio's Autonomy Stack — Suchet Bargoti, Skydio

Skydio presented its full-stack autonomy solution, demonstrating how drones are evolving from hobbyist tools into critical infrastructure. The system enables large-scale, multi-agent orchestration, allowing a single operator to manage multiple drones performing diverse tasks (e.g., utility inspection, tracking stolen vehicles) across different geographical locations simultaneously. The core technical advancements involve splitting intelligence between the edge (on-drone actions) and the cloud (long-term planning, heavy lifting), utilizing World Models for global path planning, and employing Visual Language Models (VLMs) for agentic, rule-free object tracking and semantic reasoning.

Key takeaways

1. Drones as Infrastructure 2:00

   Skydio is positioning its drones as critical infrastructure, with thousands of docks deployed across utilities, public safety, and construction sectors. This allows for continuous, reliable operation (day/night, rain/sunshine) and scales beyond the limitations of requiring a dedicated pilot for every incident.

2. Full-Stack Autonomy Architecture 18:50

   The autonomy stack splits intelligence between the edge (for immediate actions) and the cloud (for heavy lifting and long-term planning). This architecture is designed to maintain high reliability (targeting 99.9999%) while managing vast amounts of data and complex decision-making.

3. Agentic Orchestration

   The system moves beyond hand-coded rules by using agentic tools. A VLM can receive a high-level command (e.g., 'find a white Jeep') and autonomously access APIs and tools to command a drone's trajectory, enabling 'find and follow' without specific coding for every scenario.

• AI Engineer
• Drone Autonomy
• AI/ML
• Visual Language Models (VLM)
• World Models
• Edge Computing
• Cloud Computing
• Agentic Systems
• Infrastructure Technology
• Skydio
• Skydio LinkedIn

Watch on YouTube Full article

Sep 16, 2026 · 20:36

Connect AI to Billions of Legal Documents — Simon Eskildsen, turbopuffer & Jacob Lauritzen, Legora

This talk details the architectural challenges of scaling AI-powered legal search (Legora) across billions of documents while meeting stringent regulatory requirements for data residency and physical isolation. The speaker describes a critical failure point in using traditional relational/vector databases (like PostgreSQL/Elasticsearch) with aggressive partitioning, where cache thrashing caused search latency to spike from 100ms P99 to 20 seconds. The solution was adopting TurboPuffer, an object storage-native search engine. TurboPuffer solves these issues by making the 'namespace' the atomic unit of storage, allowing for per-namespace encryption and efficient handling of cold data in object storage, thereby achieving massive latency improvements and cost efficiency.

Key takeaways

1. Scaling Failure in Traditional Databases 18:20

   Initial attempts using PostgreSQL (PG vector) and aggressive partitioning (4,000 partitions) failed at scale because mixing hot and cold projects caused the system to thrash the cache, spiking P99 latency from 100 milliseconds to 20 seconds.

2. Object Storage Native Architecture 12:00

   TurboPuffer is designed to minimize round trips by writing directly to object storage (e.g., S3). This architecture is ideal for regulated environments because the namespace can be the atomic unit for encryption and physical isolation (e.g., per-bucket, per-key).

3. Handling Cold Data and Isolation 12:00

   By making the project the unit of storage (namespace), idle projects can rest in object storage, costing nothing, while still maintaining the ability to enforce physical separation and key control required by large financial institutions.

4. Search Optimization for Memory Hierarchy 16:40

   TurboPuffer optimizes search by organizing vector data into a tree structure (like a complex B-tree) rather than navigating a graph. This design minimizes expensive round trips across the memory hierarchy (DRAM $ ightarrow$ NVMe SSD $ ightarrow$ Object Storage).

• AI Engineer
• Search Engineering
• Object Storage
• Vector Databases
• Data Residency
• AI/ML
• TurboPuffer
• Legora

Watch on YouTube Full article

Aug 11, 2026 · 1:35:20

🔬Biology Is Turning Into Software — Matt McPartland & Neil Patel, Chai Discovery

The intersection of biology and software is transforming drug discovery from a slow 'waterfall' process into an agile, iterative loop. Chai Discovery leverages advanced AI models (Chai-2, Chai-3) that function as sophisticated design suites—more akin to SolidWorks or Figma than ChatGPT. These platforms enable the co-design of protein sequences and structures, allowing researchers to move beyond simple structure prediction toward generating novel therapeutic candidates with high precision for complex modalities like ADCs and bispecifics.

Key takeaways

1. Platform Design vs. Chatbot Interface

   The product is designed as a visual, highly functional design suite (like Autodesk or SolidWorks), allowing users to 'paint' epitopes and generate binders, rather than operating through conversational prompts.

2. AI Model Progression for Drug Design 42:29

   Chai models progressed from Chai-1 (structure prediction) to Chai-2 (all-atom diffusion model capable of design), crossing the threshold into generating candidate molecules that bind to a target structure, which is critical for drug development.

3. Shifting Drug Discovery Paradigms 20:39

   The process is moving from a costly, multi-year 'waterfall' model (target discovery $ ightarrow$ hit discovery $ ightarrow$ optimization) toward an agile, iterative loop where AI models provide rapid, promising candidates for continuous refinement.

• Latent Space
• AI/ML
• Protein Engineering
• Drug Discovery
• Computational Biology
• Software Architecture
• Chai Discovery
• Eli Lilly
• Pfizer
• Novartis
• argenx

Watch on YouTube Full article

Aug 4, 2026 · 13:55

What Are Large Database Models? AI for SQL Data

Large Database Models (LDMs) represent a significant advancement in applying AI to enterprise data by bringing semantic capabilities directly into SQL and relational databases. Unlike Large Language Models (LLMs), which train on general text, LDMs are trained specifically on selected tables or views within a structured database. This allows organizations to unlock the estimated 99% of critical business data—often locked behind encryption and access controls—without needing to move it.

Key takeaways

1. LDM Functionality vs. Traditional SQL 2:15

   Traditional methods require data scientists to manually write rigid SQL filters (e.g., `where age is between 20 and 40`) and move data to an analytics platform, which is slow and expensive. LDMs use vector representations learned from co-occurring values across columns to perform semantic queries, eliminating the need for manual field selection or guessing constraints.

2. Core LDM Capabilities 3:30

   LDMs enable advanced querying capabilities such as finding customer similarity (finding customers 'most similar' to a given ID), identifying unusual transactions (fraud detection), and exploring product relationships, all executed via standard SQL against the database itself.

3. Commercial Availability 9:00

   IBM launched the first LDM-based database product, 'SQL Data Insights,' which ships as part of DB2 for ZOS. A follow-up version, 'SQL Data Insights Pro,' extends this approach to unstructured text and adds incremental model refresh.

• IBM Technology
• Large Database Models
• AI/ML
• SQL
• Data Engineering
• Vector Embeddings
• Relational Databases (RDBMS)
• Large Database Models (LDMs)
• IBM AI Newsletter

Watch on YouTube Full article