Lei Xu

New feature for Ray users! We just launched the persistent Ray Task Dashboard on Anyscale,  making it easier than ever to monitor, debug, and optimize Ray workloads at scale. ✅ Logs and observability that persists beyond cluster shutdown ✅ No more 10k task ceiling ✅ Built on the new Ray Event Export framework - available from Ray 2.49 Say goodbye to ephemeral dashboards and hello to always-on observability. Read More: https://lnkd.in/etqvDb4K

40

1 Comment

🧠 How JFrog ML Is Transforming the AI Software Supply Chain AI is becoming core to software delivery, but most ML pipelines remain fragmented, insecure and hard to govern. JFrog ML extends JFrog’s Software Supply Chain Platform to unify DevOps, DevSecOps and MLOps all in one secure enterprise-grade ecosystem. JFrog ML delivers:  ✅ End-to-end model lifecycle management ✅ Built-in versioning, metadata and promotion workflows ✅ Continuous security scanning of AI models and dependencies ✅ Seamless CI/CD integration for ML artifacts ✅ Multi-cloud and hybrid support with full traceability As AI scales, so do the risks. JFrog ML helps teams manage models with the same trust, traceability and automation as software, at an enterprise scale. 👉 Read the full blog here: https://lnkd.in/gkPGEfnP #MLOps #DevSecOps #SoftwareSupplyChain #AI #JFrog

26

2 Comments

LiteLLM is an open-source proxy that connects your app to LLMs through one unified interface. Paired with Redis, it gives AI and ML teams a simple yet powerful way to unify access to LLMs, accelerate response times, and make AI apps real-time. Redis handles performance, memory, and data coordination that modern AI apps demand, while LiteLLM handles abstraction and routing. Our AI Product Marketing Manager Rini Vasan dives into: ▶️ What LiteLLM is ▶️ Why it works well with Redis ▶️ How LiteLLM and Redis work together Learn how to scale your LLM gateway with LiteLLM & Redis: https://lnkd.in/gqsTebF7

32

In this post, we demonstrate how to integrate Amazon SageMaker HyperPod with Anyscale platform to address critical infrastructure challenges in building and deploying large-scale AI models. The combined solution provides robust infrastructure for...

1

“1 million rows per second.” In his opening keynote, our CEO Max Liu walked through what that really means in production: - TiDB processing 1M rows/sec - 10x higher throughput for cross-region replication - A data layer built to make multi-tenant architectures actually practical at scale If you’re thinking about SaaS, multi-region, or AI-heavy workloads, this is the kind of infrastructure ceiling you want. 🎥 Watch the full keynote from Max Liu at #TiDBSCaiLE to see what’s behind the numbers: https://ow.ly/BgKl50YetBq #TiDB #DistributedSQL #MultiTenancy #CloudNative #Scalability

25

Yasmin COO at StormForge discusses the common pitfall of overprovisioning in Kubernetes clusters. She reveals how teams blindly set their HorizontalPodAutoscaler target utilization to 60% simply because it appears in the Kubernetes documentation, effectively wasting 40% of CPU resources. She outlines two optimization strategies: monitoring static applications through Grafana dashboards for accurate resource allocation and implementing automated scaling for dynamic workloads with varying resource needs throughout different time periods. Watch the full interview: https://ku.bz/Bs49B8sYG This interview is a reaction to Alexandre's episode https://ku.bz/z2Vj9PBYh

32

2 Comments

Build a real-time lakehouse with Apache Doris + AWS Glue + S3 Tables ➡️ Join our meetup in SF to see the full demo: https://lnkd.in/gRmeZWUd 🔧 Stack: - S3 stores Iceberg table data - AWS Glue as a central metadata catalog for schema and table management - Apache Doris (via VeloDB) brings unified batch + real-time SQL analytics on top 💡 Pros of this setup: - Real-time analytics: Doris delivers sub-second, high-concurrency queries directly on S3 data. - Open architecture: Keep data in open formats (Parquet, Iceberg) for multi-engines access. - Simple AWS integration: Glue makes tables easily managable across AWS services (Athena, EMR, Redshift Spectrum...) - Cost-efficient: Cheap S3 storage, elastic Doris compute, and no complex ETL pipelines. Our quick demo below shows connecting Glue and S3 to Doris through VeloDB Studio, then create and query a new table, all in real time. 🎟️ RSVP link at the top, join our event in SF to see our demo and demos from Uber, Pinterest, and Datastrato. #AWS, #ApacheDoris, #Gravitino

13

Bi-directional replication between RDS and AlloyDB—done right. In our latest Engineering Blog, Zencore Senior Cloud Architect Diego Sucaria shares a detailed technical guide to implementing pglogical-based replication between Amazon RDS and Google Cloud AlloyDB. The post covers: • Node configuration • Subscription management • Sequence handling • Key failure scenarios that can break replication or crash instances Ideal for teams managing hybrid or multi-cloud Postgres environments. Read the full guide: https://lnkd.in/eEyVfhre #Zencore #GoogleCloud #AWS #AlloyDB #AmazonRDS #PostgreSQL #DataReplication #CloudArchitecture #MultiCloud #EngineeringBlog

19

🚀 Announcing state-machine-amz-go v1.0.6! (major update) I’m excited to share a major milestone for my open-source AWS Step Functions–compatible state machine library in Go: v1.0.6 is now live! This release transforms the library from a lightweight executor into a production-ready workflow orchestration engine with full persistence, execution history, and multi-backend support—all while keeping the v1 API 100% backward compatible. ✅ Persistent State Machines – Automatic state saving, retries, and audit trails ✅ Dual PostgreSQL Strategies:  • GORM (rapid dev, auto-migrations, built-in stats)  • Raw SQL (max performance, fine-tuned queries) ✅ In-Memory Mode – Perfect for testing ✅ Rich Query API – Filter, paginate, and analyze executions ✅ Execution Statistics – P50/P95 durations, success rates, and more (GORM only) Under the hood: Clean separation of concerns (stateless core + pluggable persistence) Full ACID-compliant history tracking Optimized indexes & JSONB support Zero breaking changes for existing users Whether you're running simple workflows or building a mission-critical orchestration layer, v2.0 gives you the flexibility, observability, and reliability you need—without vendor lock-in. 🔗 GitHub: https://lnkd.in/dtzeQuSR 🔗 GoPkg: https://lnkd.in/dc2Ft6VA 📘 Full Release Notes: See the detailed changelog (link in comments!) Huge thanks to everyone who’s used, starred, or contributed! 🙏 Next up: Redis & DynamoDB backends, Prometheus metrics, and a web dashboard 🚀 #Go #Golang #StateMachines #WorkflowOrchestration #PostgreSQL #GORM #OpenSource #CloudNative #BackendEngineering #DevTools #AWSStepFunctions #DistributedSystems

18

3 Comments

LLMs look impressive in demos — but become inconsistent once they're embedded in real workflows. BASE Studio helps teams fine-tune models from SFT to RLHF using data built for real-world use, not toy examples: 🔵 Multi-turn conversations that preserve context 🔵 Clear, followable instructions models can reliably execute 🔵 Code data that explains *why* decisions are made — not just the final answer 🔵 Reasoning data that strengthens step-by-step thinking For multimodal models, we also align prompts with visual inputs to reduce inconsistency in text-to-image and text-to-video generation. The result? Models that behave more consistently — even after deployment. Learn more 👉 https://boden.ai #EmbodiedAI #Robotics #AITraining #TrainingData #MachineLearning #DataPipeline #RobotLearning #MultimodalAI #Automation #ComputerVision #AIInfrastructure #Annotation

1

Keeping your #software supply chain safe starts with what you bring in. JFrog Curation helps teams automatically vet open-source packages before they enter your environment—catching malicious, vulnerable, or non-compliant components early. The result? Cleaner builds, fewer surprises, and more confidence in every release. #OpenSource

2

The world is compute-hungry & rightly so. Aethir’s distributed GPU network lets anyone become a Cloud Host, monetizing high-performance GPUs in exchange for stable ATH rewards. Key Benefits of Becoming an Aethir Cloud Host: 🔹 High GPU utilization (95%+) for maximizing compute revenue. 🔹 Anyone with high-performance GPUs can earn stable revenue in ATH tokens. 🔹 Access to 150+ enterprise clients across AI, gaming, and Web3 sectors. 🔹 Transparent revenue model based on GPU quality, usage, and uptime. 🔹 Dedicated 24/7 support. 🔹 Simple and quick onboarding process. But how do you become a Cloud Host? Our blog is a step-by-step guide ⬇️ https://lnkd.in/dy_nAy46

11

If you're building scalable applications, you need to master indexing. Our sister site, roadmap.sh, just launched a comprehensive Elastic search Roadmap that walks you through everything from data modeling to high-throughput search queries. https://lnkd.in/g3Mr8_yV

1

Please Note: 1. The MVP for Fast Microservice Architecture for Reducing Large Language Model Hallucinations we had demo'ed along with a preprint on Zenodo has been renamed for technical clarity and to disambiguate from 3P websites that have cropped up. The updated work can be found at https://lnkd.in/dKmn7Sza 2. Decided to rename all work except game AccurateMisplaced with non product names for now - until release . This means code name for Hydra Auth may also change to be more meaningful without attempting to reserve any product names as it is unnecessary right now. 3. Best place to track all updates is at company website https://gigi-sehgal.com #GigiSehgalLLC #TopicAwareInferenceBoost #Auth #Cybersecurity

1