What shipped in AI engineering: November 2026 (so far)
339 posts from 14 sources, grouped by area and ranked by relevance to engineers shipping production AI.
Models & LLMs
Model releases, evaluations and LLM research · 70
- Q&A: Enabling more human-like interactions and understanding in multimodal foundation models
Researchers from MIT‑IBM have introduced data‑efficient benchmarks and techniques to surface latent multimodal knowledge in foundation models without heavy annotation.
- NVIDIA Nemotron 3.5 Super VL is now available in Amazon SageMaker JumpStart
NVIDIA Nemotron 3.5 Super VL, a 120 B‑parameter vision‑language model with a 12 B‑active‑parameter MoE language backbone and C‑RADIO vision encoder, is now available through Amazon SageMaker JumpStart.
- How Stretto built an AI voice agent with Amazon Nova 2 Sonic
Stretto created “Tara,” an AI voice agent that replaces web‑form bankruptcy intake with a phone‑based, speech‑to‑speech interview powered by Amazon Nova 2 Sonic, a foundation model available through Amazon Bedrock.
- Anthropic launches Fable 5.2 and Mythos 5.2 with science, coding upgrades
Anthropic announced the release of Claude Fable 5.2 and Claude Mythos 5.2, its two most capable large language models to date, sharing identical core capabilities.
- Introducing Gemini 4 Neon
Gemini 4 Neon is a new Google DeepMind model that delivers frontier‑level intelligence for complex software engineering, knowledge work, and scientific workflows, featuring a 1 million‑token context window and pricing of $3 / M input…
- Welcome Inkling 2 Mini, Cutting-edge Performance for your Local Setup
Inkling 2 Mini is a 26 B‑parameter multimodal Mixture‑of‑Experts transformer that achieves the capacity of a 48‑layer model by looping 12 middle decoder layers three times, resulting in 3 B active parameters per token and a 1 M‑token…
Show 64 more in Models & LLMs
- How Indegene is modernizing and scaling enterprise data quality in life sciences
- Introducing the Amazon SageMaker JumpStart Model Leaderboard
- PolyAI Dialog-RSN-1 is now on Amazon SageMaker JumpStart
- Reason over time-series data with OpenTSLM on Amazon SageMaker JumpStart
- Google’s latest midtier model Gemini 4 Neon is three times faster than Carbon and wins on some metrics
- How Social Department built culture-aware video discovery
- Anthropic researcher claims to have proven Riemann hypothesis using Claude
- Clairic Hit 98% Accuracy on 72.8% of Questions via Confidence Scores
- AI helps researchers investigate how thousands of medical studies share code
- StateBench: Testing how language models track state changes over long documents
- Token Pooling-Aware Training for Multi-Vector Embedding Models
- Visualizing Information Entropy in Language Models
- Bring your business context to ChatGPT, Claude, Microsoft Copilot Cowork and Grok Bot
- DaisyRose-27B: Pushing the Frontier of Post-Training Efficiency
- Genie 4 is now available
- Thinking Machines debuts Inkling 2 Small amid reported $5B fundraising push
- Google adds computer use capabilities to Gemini’s Mac app
- Introducing Unity Gateway Insights
- Understanding Positional Embeddings in Modern LLMs
- Helping AI text detectors keep up with changing conditions
- OpenAI rolls out GPT-6.2 Astra for professional work and GPT-6.2 Sol for everyday work
- Alibaba launches Qwen4-Max with training, attention optimizations
- Nemotron-Personas-Finance-Japan: A Synthetic Dataset Capturing Japan's Demographics and Financial Context
- NVIDIA Sweeps MLPerf Training and Endpoints as Vera Rubin Debuts With Leading Performance
- BotPsy: On-Device Personal Assistant with a Sub-Billion-Parameter Agentic Language Model
- After 10 Heads in a Row, Small LLMs Bet Up to 82% on an 11th
- How to Check Local LLM Output with Ollama, Gemma 4, and Nimble
- Change One Monty Hall Rule and Five LLMs Still Say Switch
- Google enters decision model market with Gemini Decisions API
- At Ignite 2026, Microsoft reinvents Copilot as the autonomous engine of enterprise work
- llama.cpp v1.0.0 : A stronger foundation for the local AI ecosystem
- LLMON: An LLM-Native Markup Language, now in Granite 4.1 3B and 8B
- Introducing Gemini 4 Hydrogen
- Gemini 4 Carbon is now available
- Opening the datasets behind our AI weather models
- Decision Models in Hugging Face
- Palette: teaching a small model to build decks in your design language
- The Complete Guide to LLM Evaluation
- Google releases Gemini 4 Hydrogen, a lower-cost model said to outscore Argon
- What Are Diffusion Language Models, and Why Are They Different From GPT?
- DiffusionGemma: Hands-On with Google’s Text Model That Writes All at Once
- Mitra-v2: Scaling synthetic pretraining for tabular foundation models
- Reflection's Beam Is the New Open-Weight Model Everyone Is Talking About
- How to Validate LLM Output in a DuckDB and dbt Pipeline
- The Wizard of Open Weights: Chris Alexiuk Makes AI Easy With Open Models
- Mirendil raises $1B in funding to speed up AI model development
- Quantile Regression Caught the House Price OLS Missed by $55K
- XYBench: Helping with the problem behind the question
- Using wearables to enable cardiovascular-kidney-metabolic (CKM) risk profiling for personalized coaching
- Your Experts Have Partners. There's HOPE for Keeping Them Together
- Detecting signs of hypertension using wearable sensors
- Dharma Field Extractor: Any Document In, Structured Fields in JSON Out
- How to Train Your own Decision Model like Jev using Unsloth
- Tencent releases Hy4, a powerful frontier AI model introducing visual understanding and computer use
- Anthropic launches IPO roadshow with $64B fundraising goal
- Getting Reliable Structured Outputs from LLMs: Validation, Retries and Constrained Decoding
- Why a Tiny GPT Can Know Two Facts and Still Not Connect Them
- Real-time fraud detection with Feature Views
- MIT-led team receives funding to develop next-generation alternatives to modern weed killers
- Hamed Karimi wins ECML PKDD Test-of-Time Award
- Clearing up a robot’s fuzzy memory
- Regina Barzilay and Jim Collins win Pritzker Prize for AI and Science Research Excellence for work in AI-enabled antibiotic discovery
- Attention Variants and Memory Efficiency
- Financial institutions confront the complexities of scaling AI
Agents & MCP
Agent frameworks, tool use, MCP and orchestration · 108
- The agent evaluation playbook: Built-in, DeepEval, and AutoEval on Amazon Bedrock AgentCore Evaluations
The article outlines how to evaluate Amazon Bedrock agents using AgentCore’s built‑in evaluators (e.g., Builtin.GoalSuccessRate, Builtin.Helpfulness) together with third‑party DeepEval and AutoEval evaluators hosted in managed mode.
- ServiceNow on-behalf-of token exchange with Bedrock AgentCore Gateway
This article details how to configure Amazon Bedrock AgentCore Gateway to perform an on‑behalf‑of (OBO) token exchange with Microsoft Entra ID, enabling an AI agent to call the ServiceNow REST API while preserving the original user’s…
- Towards a Science of Capable Autonomous Agents & Provable Safety — Part II
Part II examines how to obtain provable safety guarantees for increasingly capable autonomous agents, emphasizing that higher capability does not inherently ensure safety.
- Let Decision Models Watch Your Agents
This article presents a hands‑on guide for detecting behavioral anomalies by monitoring an agent’s evolving execution history with decision models.
- Introducing harness adapters: adapt your favorite model to your harness to improve performance and cost
The authors demonstrate that training a custom RL‑trained LoRA adapter (“harness adapter”) for Qwen3.8‑27B inside specific agent harnesses (Pi, OpenCode, Hermes Agent) raises Terminal‑Bench 2.1 pass@1 from 39 % to 55 % while cutting…
- Into the Omniverse: How Developers Turn Ideas Into Simulations With Frontier AI Agents
NVIDIA developers are leveraging frontier AI agents such as GPT‑6 Astra together with NVIDIA Omniverse libraries to automate the creation and scaling of physics‑rich simulations.
Show 102 more in Agents & MCP
- Reducing token costs and latency in Amazon Bedrock AgentCore JSON workloads
- How CrossnoKaye grounds physical AI with Amazon Bedrock AgentCore
- Taxbit Assist: A grounded, role-aware AI agent on Amazon Bedrock AgentCore
- Responsible AI for agentic AI: A phased roadmap for platform and application teams
- Taxonomy of agentic primitives in Amazon Quick
- Govern AI agent access with Okta Cross App Access and Amazon Bedrock AgentCore – Part 1
- Pluggable Agentic AI Framework: A seven-layer composable enterprise agentic AI reference architecture on Amazon Bedrock AgentCore
- Amazon Bedrock AgentCore harness adds Strands harness defaults, built-in tools, and lifecycle hooks
- Protect agent tool calls with Skyflow and Amazon Bedrock AgentCore Gateway
- Build WhatsApp AI agents with Amazon Bedrock AgentCore
- How LoanPro built an agentic AI loan servicing assistant with Amazon Bedrock AgentCore
- Enable AI agents to pay for AWS Marketplace x402 products with Amazon Bedrock AgentCore payments
- Agentic engineering intelligence with Amazon Nova Multimodal Embeddings and Amazon Bedrock AgentCore
- Google upgrades Gemini’s agentic Task mode with new model, e-commerce features
- Private certificate authority support in Amazon Bedrock AgentCore Identity
- From learning to leaderboard: How Cisco upskilled 1,200 engineers in agentic AI with AWS AI League
- Own, don’t rent: Why agentic AI is pulling cloud computing back on-prem
- Taxonomy of multi-agent design patterns, anti-patterns, and lessons from manufacturing
- Automating AWS Agent Registry updates with push-based architecture
- Govern and discover records with common metadata in AWS Agent Registry
- Secure OAuth patterns for Snowflake Cortex AI through Amazon Bedrock AgentCore
- How MoEngage built agentic customer engagement on Amazon Bedrock AgentCore
- Build memory-aware AI agents with Agentic Retrieve on Amazon Bedrock Managed Knowledge Base
- Build a daily dependency watcher with LangGraph and Amazon Bedrock AgentCore web search
- How TCS built an agentic test orchestrator with Amazon Bedrock AgentCore
- Genie ZeroOps, now in public beta: An autonomous agent that fixes data pipelines, with lessons from 80,000 customer issues
- Snowflake expands agentic AI platform with business context and governance tools
- With Agentic Studio, LinearB provides the critical context needed to accelerate software delivery
- How to Oversee AI Agents at Scale With a Three-Tier Architecture
- Microsoft rolls out new first-party models and updates to agentic AI continuous improvement
- How AI Agents Become a Tool for Attackers, and How to Stop It
- How skills and agents work in AWS Quick
- NVIDIA CUDA-X Libraries Give Agents the Tools to Accelerate Scientific Discovery
- Adobe expands AI assistant into creative workflows
- Atomicwork launches certified AI coworker marketplace for enterprise IT and security
- Ian Buck Kicks Off NVIDIA’s Week at SC26 as Science Enters the Age of Agents
- Neurosymbolic Agents Explained, With a Z3 Example in Python
- Serverless MCP server for AI agent data lake access on Amazon ECS Fargate
- Use decision models with Amazon Bedrock AgentCore Evaluations
- Mathematical proof, powered by AI. And no hallucinations!
- Dell puts a single gateway between enterprise AI agents and the data they crave
- Turn buying signals into pipeline with Sendoso gifts in Genie One
- Build efficient agents with knowledge distillation on Amazon SageMaker AI
- Build composable, interoperable multi-agent systems with A2A on AWS
- From prototype to production: Engineering agents with Agent DLC
- SuperDial revamps its healthcare voice agent platform to stop AI bots wasting time calling each other
- How to Govern Coding Agent Skills Across Repos With Plugins
- Run agents securely and efficiently with the next generation AgentCore runtime
- Superhuman unveils personal AI assistant and expands Go across entire suite
- How to Design MCP Servers That Leave Orchestration to the Host
- Extend Claude Code to remote sessions on Amazon Bedrock AgentCore with security controls
- How Natera designed an agentic AI system for lab ops with Amazon Bedrock AgentCore
- Starburst adds context support to give AI agents consistent access to business data definitions
- How to Improve Coding Agent Efficiency by 30% with an Orchestrator
- Build autonomous AI agents with Amazon Bedrock and Spring AI
- Google ADK Agents That Pause, Resume, and Wait Hours for a Human
- Keyword Search Tied an 8B Embedding Model at Picking Agent Tools
- Nvidia at SC26: Agents are the new scientists, and they need a lab
- A Practitioner's Guide to 6 Agentic GraphRAG Patterns
- Kubernetes gets a second act as AI agents turn one prompt into thousands of inference calls
- How to Build Multi-Agent Systems with the OpenAI Agents SDK
- How Block powers collaborative AI agents with Buzz and Unity Gateway
- Vera Delivers the Best of Both Worlds: Highest Throughput and Fastest Agents
- Meta introduces Muse Ember to power cost-efficient AI agents
- How Greenlight Financial built an autonomous SOC solution on Amazon Bedrock AgentCore
- Humyn Labs raises $47M to bridge the gap between AI training and real-world robotics
- Getting started with Web Answer on Amazon Bedrock AgentCore
- Gemini Robotics 2.5 accelerates real-world general dexterity
- Four layers of context engineering for production AI agents: How Deep Insight applies them in working code
- Data From 100+ AI Agents Shows They Fail by Silent Drift, Not Errors
- AI Day Seoul Brings Together Korea’s Industry Leaders, Startups and Developer Communities
- AT&T uses agentic AI to reinvent business processes from the ground up
- ToyBlocks: A controlled world for robot learning and evaluation
- Agentic cheminformatics with Google Antigravity
- CWE-bench v2: More coverage, quality, and transparency
- Marble Cognition raises $3.75M for its AI agent workers that watch and learn how people work
- AI agent tooling provider LangChain secures $300M investment
- What AI Agents Speed Up in PhD Research, and What They Can't
- Three Techniques to Understand Your Codebase When Coding Agents Write the Code
- The Enterprise Context Layer: A Practitioner's Guide to 6 Advanced Architectural Patterns
- Korea’s AI Ecosystem Advances and Secures Agentic AI With NVIDIA
- Give Your Python AI Agent Persistent Memory with One JSON File
- NiCE’s next-generation Cognigy bets the hard part of AI agents is running them, not building them
- Agents idling while they wait on humans push Kubernetes to hand back unused resources
- Cities Worldwide Build and Scale AI Services With NVIDIA Physical and Agentic AI
- AI Assistant vs. Copilot vs. Agent, and When to Use Each
- Red Hat accelerates agentic adoption across hybrid environments
- Agile Deliberation: How AI helps humans define the gray areas
- Bring structured and unstructured data together with Amazon Quick
- Google retools Kubernetes for AI agents instead of building a new stack
- Advancing science & engineering with Antigravity’s /teamwork
- The AI Agent Memory Stack: Building Blocks, Patterns, and Frameworks
- As AI agents take control, Nvidia rethinks the security playbook
- Sandboxing AI Agent Code Execution in Python: Containers, MicroVMs and Beyond
- Agents may soon be the main workload, and most of the time they’re doing nothing
- Same model, better answer: Enterprise data is what turns generic AI into a business tool
- AI Helps Recyclers Turn Mixed Waste Into Reusable Material
- NVIDIA Expands Open Source Commitment With OpenShell Contribution and New AI Infrastructure
- Practice Makes Precise: At CoRL, NVIDIA and Partners Reveal How Robots Rehearse Delicate Work in Simulation
- AI-assisted learning platform helps robots reason beyond literal commands
- Oracle wants to pull stalled AI agents out of demo mode and into production
- The Pulse: Details of incident.io’s AI spend
Inference & infrastructure
Serving, latency, GPUs and deployment · 98
- Amazon SageMaker AI now detects latency drift, idle GPUs, and missed optimizations on your inference endpoints
Amazon SageMaker AI now automatically monitors generative AI inference endpoints (vLLM containers on GPU instances) for latency drift, idle GPU capacity, and missing optimizations, surfacing findings via AWS Health.
- How Salesforce is advancing global-scale model serving with cross-Region inference routing in Amazon SageMaker AI
Salesforce’s Agentforce Foundation team solved capacity fragmentation across 30+ global AWS Regions by deploying a managed, capacity‑aware cross‑Region inference router as a regional SageMaker AI resource.
- 180B Parameters, 32GB of RAM: Pushing MoE Inference to the Limit with Transformers
The article demonstrates that the Strata inference engine can run a 180 B parameter Mixture‑of‑Experts (MoE) model by streaming experts from SSD, achieving token‑generation quality within 6 % of the on‑GPU Strata baseline.
- Model Compression Beyond Quantization
This article outlines three core model compression techniques—pruning and sparsity, structured pruning, and knowledge distillation—each targeting parameter reduction to improve efficiency.
- Accelerate GPU training on Unity Catalog tables with UCDataset
The article announces UCDataset, a feature that enables faster GPU‑based training directly on Unity Catalog tables, reducing data movement overhead.
- Why your fine-tuning job takes longer than the FLOPs say
Predicting end‑to‑end fine‑tuning time with a simple FLOP‑based formula explains little of the observed variance (R² = 0.31, MAE = 12.41 h) on 272 production SFT jobs on NVIDIA H200 GPUs.
Show 92 more in Inference & infrastructure
- Exa boosts embedding throughput by 50% to index billions of documents on AWS AI chips
- Amazon SageMaker Inference: 2026 launches in review, part 2
- Introducing Amazon SageMaker HyperPod Global Inference Gateway: One endpoint across Regions and clusters
- Build a semantic layer for agentic AI with EDB Postgres AI and Amazon Bedrock AgentCore
- Full Fine-Tuning (FFT) on SageMaker AI Serverless Model Customization
- The attendee’s guide to the AWS re:Invent 2026 AI/ML track
- How Standard Bank Group automates AI risk assessment for generative AI
- Introducing enhanced infrastructure observability in Amazon SageMaker HyperPod
- Keep inference running through regional capacity shortfalls with cross-Region inference routing in Amazon SageMaker AI
- Using NVIDIA Run:ai GPU Fractioning on Amazon EKS
- Managed NVIDIA Dynamo on Amazon SageMaker HyperPod Inference
- Route coding agent requests to the right model on Amazon Bedrock with NVIDIA NeMo Switchyard
- Time-Series Models in the Streamhouse: A New Frontier and Use Cases
- Introducing usage-based pricing for Amazon SageMaker AI bidirectional streaming
- Cut generative AI costs: Code for rules, models for judgment
- Semantic diffing of formal policies with anti-unification
- NVIDIA and Microsoft Expand Accelerated Computing From Azure to Windows PCs
- Rumble-2-Flash: an open 0.5B speech model for 40+ languages
- SageMaker Inference now supports Multi-Instance GPU to maximize GPU utilization for model endpoints
- Enterprise AI is pushing cloud infrastructure beyond the GPU
- D-Matrix bets stacked DRAM can solve inference’s memory bandwidth problem
- Hitachi Vantara bets smarter storage can close AI’s pilot-to-production gap
- Study reveals data centers’ long-distance damage to air quality
- NVIDIA RTX Spark Compact Desktops Out Now
- Amazon SageMaker AI real-time endpoints now support KV cache aware routing
- AI and supercomputing merge as universities build GPU clusters at cloud-provider scale
- Getting AI out of the demo and into businesses globally
- Generative media inference startup fal raises $1.3B at a $16B valuation
- Introducing workload recovery and expanded infrastructure checks on Amazon SageMaker HyperPod
- Detecting and recovering from silent data corruption on Amazon SageMaker HyperPod
- The Pulse: What can we learn from the Railway hack?
- New NVIDIA Cosmos 3 Open Models for Interactive World Simulation
- Three insights you might have missed from theCUBE’s coverage of Dell’s “AI Leadership Symposium”
- NVIDIA Announces Rubin HPC — A New GPU to Accelerate Agents for Science, Engineering and Analytics
- How AUMOVIO centralized AI access with Databricks Unity Gateway
- Build a model improvement flywheel with Amazon SageMaker HyperPod Data Capture
- Cost-optimizing AssemblyAI speech-to-text inference on AWS with Amazon EC2 Spot
- Deploy frontier AI models on AWS Trainium with SGLang-Neuron
- Google Cloud brings AI agents to the supercomputing bottleneck
- AI’s GPU focus meets a CPU reality check at Dell and AMD
- How we raised Databricks cluster rate limits by 7.5×
- Activation and Floating-Point Quantization
- Architecture patterns for fast session startup on Amazon Bedrock AgentCore runtime
- Production MLOps at GS Caltex with Amazon SageMaker Canvas, part 1
- CUDA at 20: How the Platform Transformed Scientific Computing
- OpenFold3 on Amazon SageMaker JumpStart: Predict biomolecular structures
- Dell bets liquid-cooled, software-defined storage can keep AI factories fed
- Hugging Face and OpenRouter join forces to make open-weight models easy to run
- Sovereignty is the AI story everyone mentions and few have planned for ahead of 2027
- Idle, then spiking: Why AI agents are breaking Kubernetes resource planning
- AI’s hidden energy tax: More than 80% of scale-up traffic is just memory moving around
- The Definitive Fine-Tuning Guide for 3D Generation Models
- Transformers Model Integration Contracts
- Serving Scikit-LLM Pipelines as REST APIs with FastAPI
- Claude Platform now supports native OpenTelemetry observability in Databricks MLflow
- Lambda Boosts AI Factory Resiliency and Efficiency With NVIDIA Photonics
- Building ultra-low-latency voice AI applications on AWS
- Benchmarking Scikit-Ollama vs. Groq-Backed Scikit-LLM: Speed, Cost, and Accuracy
- Continuous Batching Beats Static and Dynamic by Up to 2.91× in vLLM
- NVIDIA Expands Slurm to Bring Quantum Computing Into Supercomputing
- Five months to one week: How Texas A&M is speeding scientific research
- NVIDIA DGX Spark Developer Kit Turns a Desk Into a Supercomputing Lab
- How the World’s Most Advanced AI Infrastructure Supply Chain Invents the Future
- At SC26, Nvidia debuts new Rubin HPC chip, scientific research libraries
- Qvyon launches with $8M to improve safety for physical AI deployments
- Gemini Omni 1.2 Flash: Fast and cost-effective with longer clips
- A Closer Look at Phonon TTS. Launching High-Quality On-Device Speech That Tops the Benchmarks.
- How H6.ai builds AI-powered financial automation with Lakebase Postgres
- How to Use Apple's fm CLI to Run On-Device LLMs From Python
- NVIDIA Dynamo Recipes Bring Korea’s Leading Open Models to Production
- Floating-Point Precision and Its Effects to Model Inference
- Choosing an LLM Inference Optimization Technique: A Decision-Tree Approach
- Weight-Only Quantization
- New research could lead to faster, trustworthy AI models for climate prediction
- More Performance From Every Byte: How Full-Stack Codesign Solves AI’s Memory Challenge
- Build a Robot Pick-and-Place Simulation with SimReady Assets
- Google’s GKE Labs turns reinforcement learning into a Kubernetes workload
- AlphaGenome 2: Deciphering how cells read DNA to better understand disease
- Egyptian and Nigerian English join the Open ASR Leaderboard
- From Chips to Skills, NVIDIA and Korea Expand Sovereign AI Collaboration
- NVIDIA Announces New Planned Investments in Korean Startups at the MSIT-NVIDIA AI Startup Partnership Forum
- AI Is Helping Care Teams Get Patients the Right Care, Sooner
- Korea Advances Physical AI Across Industries With NVIDIA
- How NVIDIA Open Source Tools Are Helping Researchers Build Next-Generation 6G Networks
- ‘Hogwarts Legacy’ Gets a Magical NVIDIA DLSS 5 Upgrade on GeForce NOW
- AI Is Changing How Semiconductor Fabs Work — and Engineers Are Moving Faster Because of It
- How Jev Makes Decisions About Images, With 4 Python Examples
- What to expect during SC26: Join theCUBE Nov. 17–19
- Fast Attention Kernels and Operator Fusion
- CNCF data shows cloud native is the on-ramp to AI, and maintainers are paying the price
- Rack-scale AI: How Dell and AMD are building infrastructure
- University of Utah retains operational sovereignty as Redtail expands research access
RAG & retrieval
Retrieval, vector search and knowledge bases · 6
- Accelerating generative AI delivery with reusable assets
The AWS Generative AI Innovation Center now offers a catalog of 23 production‑ready, open‑source reusable assets—including RAG pipeline components, document ingestion modules, and agent scaffolding templates—that can shrink development…
- Design responsible AI controls for Amazon Aurora PostgreSQL
This article shows how to implement the Explainability dimension of AWS Responsible AI for a Retrieval‑Augmented Generation (RAG) customer‑support chatbot built on Amazon Aurora PostgreSQL‑Compatible Edition.
- Graph-RAG for Enterprise Documents: Building a Knowledge Graph from PDFs, Wikis, and Databases
The article introduces a Graph‑RAG approach for enterprise documents, aiming to construct a knowledge graph from sources such as PDFs, wikis, and databases.
- Local LLMs Handle Textbook RAG but Fail on New Research Papers
Local large language models achieve strong Retrieval‑Augmented Generation (RAG) results on textbook material, scoring between 12 and 20 out of 20, but their performance collapses on recently published research papers, dropping to 0‑9…
- RAG Retrieval Lost 40 Points When Distracting Passages Grew to 2.68M
A reproducible evaluation shows that RAG retrieval performance can drop dramatically—by 40 points—when the corpus of distracting passages expands to 2.68 M documents.
- Prepare web content for Amazon Bedrock Knowledge Bases with Crawl4AI
Amazon Bedrock Knowledge Bases provides a fully managed RAG pipeline that can ingest web content via the built‑in Web Crawler connector, but for sites that require content pruning, targeted extraction, or custom metadata you can prepend…
Coding & dev tools
AI-assisted coding and Python tooling · 7
- Meta acquires AI coding startup Conductor, whose app runs several coding agents at once
Meta Platforms Inc. has acquired AI coding startup Conductor, whose software enables developers to run multiple coding agents concurrently, each operating in an isolated copy of the codebase.
- The 0.95 Expectile Is Not the 95th Percentile
The article clarifies that a 0.95 expectile is mathematically distinct from the 95th percentile, demonstrating the difference with exact uniform calculations and 100 synthetic Python runs.
- As AI takes over coding, Oracle rethinks how applications are built
Oracle previewed a new approach to application development that integrates AI-driven code generation, aiming to address the maintenance, governance, and understandability challenges of machine-generated software.
- Build Your First Multi-Tool AI Agent with Ollama and Python
The article demonstrates how to construct a multi‑tool AI agent in Python that leverages Ollama to orchestrate calls to three distinct database tools.
- New AI “copilot” helps engineers design better structures
MIT researchers introduced an AI “copilot” that interleaves human feedback into topology‑optimization loops, yielding structures that are easier to manufacture and exhibit higher buckling loads while incurring only a 15‑second overhead…
- Computing Effective Dimension for Quantum ML Models in PyTorch
The article introduces methods for computing the effective dimension of quantum machine‑learning models implemented in PyTorch, covering both local and global variants, Fisher information estimation, epsilon‑ball sampling, and log‑space…
Show 1 more in Coding & dev tools
Cloud platforms & enterprise
Bedrock, AWS and enterprise adoption · 50
- Use prompt caching with open-weight models on Amazon Bedrock
Amazon Bedrock now lets you reuse long prompt prefixes with open‑weight models via implicit and explicit prompt‑caching modes, cutting input‑token costs and latency.
- Amazon Bedrock tokenomics data lake
Amazon Bedrock now has a centralized observability data lake that aggregates raw invocation logs from linked accounts into a single, queryable Athena table.
- IBM builds on hybrid cloud and governance to advance its enterprise AI operating model
IBM is extending its hybrid‑cloud and governance framework to help enterprises move AI from pilot to production, addressing the gap highlighted by Deloitte where 54% of firms aim to push at least 40% of AI projects into production…
- Cisco research: Trust, not technology, is now the speed limit on enterprise AI
Cisco research highlights that trust, rather than model capability or hardware speed, has become the primary bottleneck for enterprise AI adoption.
- Four AI Architecture Patterns for Regulated Industries
The article outlines four AI architecture patterns designed for regulated industries, emphasizing that enterprise AI cannot replace all deterministic control mechanisms.
- OpenAI and Databricks: Enterprise AI on your governed data
OpenAI and Databricks are partnering to bring enterprise‑grade AI capabilities to organizations that need to operate on governed data.
Show 44 more in Cloud platforms & enterprise
- Connect Amazon Bedrock AgentCore Gateway to private APIs in a hub-and-spoke network with centralized inspection
- How DXC Technology built a multi-model AI substrate on Amazon Bedrock
- How U Mobile built agentic telco operations with Amazon Bedrock AgentCore
- How Gen Digital built Norton Revamp’s AI-curated News Engine with Amazon Bedrock AgentCore Web Search
- Use Kimi Code CLI with models on Amazon Bedrock
- How ArisGlobal scaled generative AI inference on Amazon Bedrock
- Introducing the Web Search API on Amazon Bedrock AgentCore
- How Duco transforms financial data automation with Amazon Bedrock AgentCore
- How Airbnb uses cost per task to operate coding agents on Amazon Bedrock
- Announcing AI data preparation capabilities in Amazon SageMaker: Build high-quality datasets for model development workflows
- Introducing Kimi K3.1 Pro on Amazon Bedrock
- Novo Nordisk modernizes manufacturing process documents with Amazon Bedrock AgentCore
- Autonomous treasury agents: How Trovata built an AI-native solution on Amazon Bedrock and AgentCore
- Single-shot and agentic retrieval with Amazon Bedrock Managed Knowledge Base
- Get answers from tables with tabular search in Amazon Bedrock Managed Knowledge Base
- Optimize AI economics on Amazon Bedrock with Accenture’s sequenced approach
- Your complete guide to AWS Quick at AWS re:Invent 2026
- Best practices for writing Denied Topics in Amazon Bedrock Guardrails
- Customize answers per tenant in Amazon Bedrock Managed Knowledge Base
- ICYMI: What landed for AI builders in October 2026
- Professional work deserves professional tools. Meet AWS Quick.
- Modernizing mainframe Assembler to Java using Amazon Bedrock and AWS Transform
- Introducing India-based inference for Kimi K3 on Amazon Bedrock
- Optimizing cost per task and proving agent ROI on Amazon Bedrock
- Building a self-service LLM gateway with Amazon Bedrock and LiteLLM
- Introducing Claude Fable 5.2 on AWS
- Four Amazon Bedrock inference choices that affect cost
- Tokenomics at scale: Track and cap Amazon Bedrock spend per IAM principal with CUR 2.0 and AWS Budgets
- Migrate to Amazon Bedrock Managed Knowledge Base with your coding agent
- OpenAI Codex on Amazon Bedrock
- Fine-grained access control for Amazon Bedrock Managed Knowledge Base tools
- Cost attribution and controls for AI developer and productivity tools on Amazon Bedrock
- How Circle K accelerated DevOps incident triage to under 15 minutes with Amazon Bedrock
- Govern AI agent access with Okta Cross App Access and Amazon Bedrock AgentCore – Part 2
- The Enterprise Case for Open-Weight Models on Databricks
- Noeri raises $110M to help companies build living AI systems that self-improve over time
- Genie in Slack is now Generally Available: Governed AI cowork, right where work happens
- How Amazon Marketplace evaluates content quality with Amazon SageMaker AI
- Smart Approvals in AWS Quick: Human oversight where it matters
- Enterprises bring AI to their data as privacy rules tighten and costs climb
- Pay for Hugging Face Hub with your Google Cloud Account
- How to Measure AI Adoption, Performance, and ROI With 12 KPIs
- Hatz AI raises $24M to help small businesses go AI-native through their MSPs
- AWS’ Colleen Aubrey: Stop inventing new AI metrics — it’s your business metrics that matter
Grouped by each post's main topic tag. Summaries are AI-generated from the source and always link to it.
Get next week's releases by email
No hype. The week's best AI engineering posts, every Sunday.