AI, MLOps & Robotics Newsletter #99

The latest developments in AI, MLOps, and Robotics: Deep Seek releases more open source, Amazon’s Alexa Plus, The next generation of the Phi family, GPT-4.5, Claude 3.7 Sonnet + code, Trace & Evaluate your Agents, AWS Trainium2, and much more!

??? Shoutouts

• ?? My next Fireside Chat on 3/5: LLM Observability: Tracing, Evaluations, and Real-Time Insights. RSVP to attend live and get the recording.
• ??? My next workshop 3/12: Fine-Tune, Serve & Scale LLM Workflows
• ?? Next AI book club reading: AI Book Club: AI Engineering on 3/25 - RSVP on LinkedIn or Meetup.

?? AI News & Stories:

• DeepSeek claims ‘theoretical’ profit margins of 545% - Read on TechCrunch
• All of the announcements from Amazon’s Alexa Plus event - Read on the Verge
• Unitree Robots That Dance, Fight Earn Founder Beijing’s Acclaim - Read on Bloomberg
• The Humane Ai Pin Has Already Been Brought Back to Life - Read on Wired
• DeepSeek’s Lesson: The Future Of AI Is Decentralized And Open-Source - Read on Forbes
• Empowering innovation: The next generation of the Phi family - Read on Microsoft

?? ML Developer Tools & Resources:

• Introducing GPT-4.5 - Read on OpenAI
• Claude 3.7 Sonnet and Claude Code - Read on Anthropic
• Mind the GAP: Geometry Aware Passthrough mitigates cybersickness - Read on Google Research
• DUSt3R: Geometric 3D Vision Made Easy : Explanation and Results - Read on LearnOpenCV
• Trace & Evaluate your Agent with Arize Phoenix - Read on Hugging Face
• FastRTC: The Real-Time Communication Library for Python - Read on Hugging Face
• Building Production-Ready Compound AI Applications Just Got a Lot Easier: A RAG Example - Read on Union
• Building Secure AI Systems with Union's Defense-in-Depth Approach - Read on Union
• From Experiments to Production: Building ML Pipelines with MLflow and BentoML- Read on BentoML
• Detecting Speed Limit Signs with Edge Impulse and OpenMV - Read on Edge Impulse
• How to generate industry-specific data for AI training with Labelbox - Read on Labelbox
• Memes Are the VLM Benchmark We Deserve - Read on Voxel51
• Introducing Anthropic's Transparency Hub - Read on Anthropic
• Memory and State in LLM Applications - Read on Arize
• Enhancing Multimodal AI: Bridging Audio, Text, and Vector Search - Read on Zilliz
• BlockMax WAND: How Weaviate Achieved 10x Faster Keyword Search - Read on Weaviate
• LangGraph 0.3 Release: Prebuilt Agents - Read on LangChain
• Introducing LlamaExtract: Unlocking Structured Data Extraction in Just a Few Clicks - Read on Llamaindex
• RAG-Driven Generative AI - AI Book Club Recording - Watch on YouTube
• Quickly Start Evaluating LLMs With OpenEvals - Read on Langchain
• RAG: Fundamentals, Challenges, and Advanced Techniques - Read on Label Studio
• CUDA Accelerated: How CUDA Libraries Bolster Cybersecurity With AI - Read on NVIDIA
• You Should Build a Robot (MicroPython & Microprocessors) - Watch on YouTube

?? Upcoming AI Events

?? Join me live at my next Fireside Chat on 3/5: LLM Observability: Tracing, Evaluations, and Real-Time Insights. RSVP to attend live and get the recording.

• Build Vision AI Solutions for your clients | Mar 4th - Register on LiveStorm
• LLM Observability: Tracing, Evaluations, and Real-Time Insights | Mar 5th - Register on Eventbrite
• Mastering Agentic Applications: Harness APIs and Business Functions with Arch | Mar 5th - Register on DataScienceDojo
• Online Workshop - Building Scalable Apps with Weaviate | Mar 6th - Register on Luma
• AI Ready Data: Forging the Path to Reliable and Scalable AI | Mar 12th - Register on Data Science Dojo
• Fine-Tune, Serve & Scale LLM Workflows - LLMOps Workshop | Mar 12th - RSVP on Eventbrite
• NVIDIA GTCV (San Jose) | Mar 17 - 21st - Register on NVIDIA
• Make a NotebookLM Clone with Open Weights & Models | Mar 19th - Register on Eventbrite
• AI Engineering - AI Book Club | Mar 25 - RSVP on LinkedIn
• [Seattle] Startup Summit 2025 - AI Dev Tools - RSVP on Luma
• AI Book Club: AI Agents in Action | April 29th - Register on LinkedIn

??? AI Podcasts & Interviews:

• Telenor's Kaaren Hilsen on Launching Norway’s First AI Factory - Listen on NVIDIA
• GenAI hot takes and bad use cases - Listen on Changelog
• How to Grow (and Sell) a Data Science Consultancy, with Cal Al-Dhubaib - Listen on SuperDataScience
• Accelerating AI Training and Inference with AWS Trainium2 with Ron Diamant - Listen on TWiML
• The rise of AI agents | WandB - Listen on YouTube
• Open Operator, Serverless Browsers and the Future of Computer-Using Agents - Listen on Latent Space

?? Latest AI/ML Research Papers

• Slamming: Training a Speech Language Model on One GPU in a Day - Read on arXiv
• AISafetyLab: A Comprehensive Framework for AI Safety Evaluation and Improvement - Read on arXiv
• KV-Edit: Training-Free Image Editing for Precise Background Preservation - Read on arXiv
• Emergent Misalignment: Narrow finetuning can produce broadly misaligned LLMs - Read on arXiv
• From System 1 to System 2: A Survey of Reasoning Large Language Models - Read on arXiv
• Fractal Generative Models - Read on arXiv

Hello, world! Every week, I keep track of the most exciting news, resources, events, podcasts, and papers I find around AI, MLOps, and robotics. Sometimes I sprinkle in my opinions. Let me know in the comments what I missed this week!

You Can Support This AI Newsletter By:

?? joining the Flyte MLOps Slack Group and introduce yourself!

? Give Flyte, the open-source MLOps project I work on, a GitHub star!

Connect with or follow me (Sage Elliott) for more machine learning, MLOps, and robotics content. Have a great week! ??