{"@attributes":{"version":"2.0"},"channel":{"title":"SkyPilot Blog","link":"https:\/\/blog.skypilot.co\/","description":"Recent content on SkyPilot Blog","image":{"title":"SkyPilot Blog","url":"https:\/\/blog.skypilot.co\/skypilot-social-card.png","link":"https:\/\/blog.skypilot.co\/skypilot-social-card.png"},"generator":"Hugo -- gohugo.io","language":"en-us","lastBuildDate":"Thu, 21 May 2026 09:00:00 -0700","item":[{"title":"RL Doesn't Work on Slurm","link":"https:\/\/blog.skypilot.co\/rl-doesnt-work-on-slurm\/","pubDate":"Thu, 21 May 2026 09:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/rl-doesnt-work-on-slurm\/","description":"Online reinforcement learning for LLMs breaks Slurm's batch scheduling model. We'll discuss why, and what can be done about it."},{"title":"Cache Me If You Can: Tuning Object Stores for AI","link":"https:\/\/blog.skypilot.co\/cache-tuning\/","pubDate":"Thu, 30 Apr 2026 08:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/cache-tuning\/","description":"We ran hundreds of benchmarks to tune storage systems for distributed training so you don\u2019t have to."},{"title":"GPU Compass: Navigate the GPU Frontier Across 20+ Clouds & 2K+ Offerings","link":"https:\/\/blog.skypilot.co\/gpu-compass\/","pubDate":"Tue, 21 Apr 2026 08:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/gpu-compass\/","description":"Introducing GPU Compass: One dashboard to browse, compare pricing, and launch across every GPU cloud."},{"title":"SkyPilot Agent Skill: Let Agents Manage Your GPUs","link":"https:\/\/blog.skypilot.co\/agent-skill\/","pubDate":"Thu, 09 Apr 2026 08:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/agent-skill\/","description":"With the SkyPilot Agent Skill, your AI coding agent can launch clusters, run training jobs and manage cloud resources across any infrastructure using natural language."},{"title":"Research-Driven Agents: What Happens When Your Agent Reads Before It Codes","link":"https:\/\/blog.skypilot.co\/research-driven-agents\/","pubDate":"Wed, 08 Apr 2026 12:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/research-driven-agents\/","description":"Coding agents working from code alone generate shallow hypotheses. Adding a research phase \u2014 arxiv papers, competing forks, other backends \u2014 produced 5 kernel fusions that made llama.cpp CPU inference 15% faster."},{"title":"Scaling Karpathy's Autoresearch: What Happens When the Agent Gets a GPU Cluster","link":"https:\/\/blog.skypilot.co\/scaling-autoresearch\/","pubDate":"Wed, 18 Mar 2026 12:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/scaling-autoresearch\/","description":"Karpathy's autoresearch runs one experiment at a time. We gave it access to our GPU infra and let it run experiments in parallel."},{"title":"SkyPilot Recipes: Templatize your AI Workflows","link":"https:\/\/blog.skypilot.co\/skypilot-recipes\/","pubDate":"Tue, 10 Mar 2026 08:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/skypilot-recipes\/","description":"SkyPilot Recipes let you store SkyPilot YAMLs in a shared, team-accessible registry. Launch workloads directly from the CLI without local files."},{"title":"SkyPilot Job Groups: Run RL on Heterogenous Hardware","link":"https:\/\/blog.skypilot.co\/job-groups\/","pubDate":"Mon, 02 Mar 2026 12:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/job-groups\/","description":"SkyPilot Job Groups let you define heterogeneous RL workloads in a single YAML. Run your PPO trainer on beefy H100s, rollout servers on cheap T4s, and replay buffers on high-memory CPUs, all as one managed job."},{"title":"Don't Run OpenClaw on Your Main Machine","link":"https:\/\/blog.skypilot.co\/openclaw-on-skypilot\/","pubDate":"Thu, 26 Feb 2026 12:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/openclaw-on-skypilot\/","description":"OpenClaw gives an AI agent full access to your system. Here's why you should run it on an isolated cloud VM, and how to set that up."},{"title":"SkyPilot Admin Policies: Enforce GPU Governance Without Slowing Down AI Teams","link":"https:\/\/blog.skypilot.co\/skypilot-admin-policies\/","pubDate":"Fri, 20 Feb 2026 08:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/skypilot-admin-policies\/","description":"SkyPilot Admin Policies let you enforce cost controls, security rules, and compliance requirements automatically \u2014 without slowing down your engineering team."},{"title":"Migrating from Slurm to Kubernetes","link":"https:\/\/blog.skypilot.co\/slurm-to-k8s-migration\/","pubDate":"Tue, 10 Feb 2026 07:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/slurm-to-k8s-migration\/","description":"Moving from Slurm to Kubernetes doesn't have to mean losing the workflow you know. Here's how SkyPilot brings Slurm-like simplicity to K8s."},{"title":"SkyPilot Volumes: Fast and Persistent Storage for AI Workloads","link":"https:\/\/blog.skypilot.co\/skypilot-volumes\/","pubDate":"Thu, 22 Jan 2026 07:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/skypilot-volumes\/","description":"Mount Kubernetes PVCs to your clusters for 10-100x faster data access with persistent storage that survives across job lifecycles."},{"title":"Scaling SAM3 Video Segmentation on Multiple Kubernetes clusters and Clouds with SkyPilot","link":"https:\/\/blog.skypilot.co\/skypilot-pools-sam3\/","pubDate":"Tue, 13 Jan 2026 07:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/skypilot-pools-sam3\/","description":"Run Meta's SAM3 on large video archives distributed across AWS and Kubernetes clusters with SkyPilot Pools."},{"title":"Launch AI Jobs faster with SkyPilot Templates","link":"https:\/\/blog.skypilot.co\/skypilot-templates\/","pubDate":"Fri, 19 Dec 2025 07:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/skypilot-templates\/","description":"SkyPilot now includes predefined templates to launch clusters with popular frameworks and patterns. Deploy fully configured environments without writing long YAMLs."},{"title":"Train an agent to use google search as a tool with RL","link":"https:\/\/blog.skypilot.co\/verl-tool-calling\/","pubDate":"Wed, 17 Dec 2025 07:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/verl-tool-calling\/","description":"Train a tool-calling agent with VeRL and use SkyPilot to scale it up with independent RL trainer and env rollout"},{"title":"SkyPilot 0.11: Multi-Cloud Pools for Batch Inference, Fast Managed Jobs, Enterprise-Ready at Scale, Programmability","link":"https:\/\/blog.skypilot.co\/announcing-skypilot-0.11.0\/","pubDate":"Thu, 11 Dec 2025 06:00:00 -0600","guid":"https:\/\/blog.skypilot.co\/announcing-skypilot-0.11.0\/","description":"Announcing SkyPilot 0.11 with Pools for batch inference, faster managed jobs, and enterprise-scale improvements."},{"title":"Batch Inference for Documents with DeepSeek-OCR using a Pool of Workers on any Clouds","link":"https:\/\/blog.skypilot.co\/skypilot-pools-deepseek-ocr\/","pubDate":"Tue, 02 Dec 2025 00:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/skypilot-pools-deepseek-ocr\/","description":"Scale document OCR batch inference for RAG on multiple clouds and Kubernetes clusters using SkyPilot Pool."},{"title":"Why AWS Batch Doesn't Work for Modern AI Workloads: A Technical Comparison with SkyPilot","link":"https:\/\/blog.skypilot.co\/aws-batch-vs-skypilot\/","pubDate":"Tue, 21 Oct 2025 06:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/aws-batch-vs-skypilot\/","description":"AWS Batch works well for traditional enterprise batch processing (see their case studies 1 and 2). But AI workloads have different requirements - they’re more interactive, need flexible GPU access, and benefit from simpler iteration cycles.\nIn this post, we explore how you would use AWS Batch to run AI batch inference at scale, and explain why it does not fit well with modern AI workloads. We’ll also show an alternative approach using SkyPilot."},{"title":"How to train and scale AI math\/coding agents using VeRL on any AI infra","link":"https:\/\/blog.skypilot.co\/verl-rl-training\/","pubDate":"Tue, 14 Oct 2025 07:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/verl-rl-training\/","description":"Want to train an AI agent with RL that can solve math problems or write code? This tutorial walks you through building your own math and coding agents with step-by-step examples with plenty of screenshots to help you along the way.\nWe use VERL (a production-ready training framework) to apply RL post-training for LLM and SkyPilot to run and scale the training on any of your own AI infrastructure, including Kubernetes and clouds."},{"title":"Scaling Vector Search to 1M Documents for $0.85","link":"https:\/\/blog.skypilot.co\/redisvl-skypilot\/","pubDate":"Tue, 23 Sep 2025 00:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/redisvl-skypilot\/","description":"How to build production vector search with RedisVL and SkyPilot: 1M documents indexed for $0.85, sub-100ms queries, no Kubernetes required."},{"title":"Unlocking GPU Metrics in Kubernetes with SkyPilot","link":"https:\/\/blog.skypilot.co\/gpu-metrics\/","pubDate":"Fri, 12 Sep 2025 00:10:19 -0700","guid":"https:\/\/blog.skypilot.co\/gpu-metrics\/","description":"SkyPilot now supports detailed GPU metrics across multiple Kubernetes clusters in the dashboard for better observability."},{"title":"From 1 hour to 10 minutes: How I sped up my distributed LLM training without changing the code or GPUs","link":"https:\/\/blog.skypilot.co\/gpu-distributed-training\/","pubDate":"Thu, 11 Sep 2025 07:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/gpu-distributed-training\/","description":" This page has moved. If you are not redirected automatically, click here. "},{"title":"Scaling AI Infrastructure at Abridge with SkyPilot","link":"https:\/\/blog.skypilot.co\/abridge\/","pubDate":"Thu, 04 Sep 2025 07:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/abridge\/","description":"How we transformed our fragmented multi-cloud AI infrastructure into a unified system with SkyPilot, achieving 10x faster development cycles."},{"title":"From SLURM to SkyPilot: How Avataar cut costs 11x with multi-cloud AI infrastructure","link":"https:\/\/blog.skypilot.co\/avataar\/","pubDate":"Wed, 20 Aug 2025 08:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/avataar\/","description":"Avataar's enterprise AI content platform cut costs 11x and unlocked GPU capacity by migrating from inflexible SLURM deployment to SkyPilot's multi-cloud infrastructure."},{"title":"Self-host open-source LLM agent sandbox on your own cloud","link":"https:\/\/blog.skypilot.co\/skypilot-llm-sandbox\/","pubDate":"Tue, 12 Aug 2025 07:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/skypilot-llm-sandbox\/","description":"Your AI writes code. Now what?\nIf you’re building AI agents in 2025, you probably wondered that as well. Your LLM generates some Python code that analyzes data, manipulates files, or calls APIs. But where does it run? Most people either pay for managed execution services where they don’t control the data flow, or just YOLO it and run everything locally.\nNeither option scales, and both come with problems you probably haven’t thought about yet."},{"title":"Slurm vs K8s for AI Infra: Academic HPC vs Cloud-Native Reality - the non-ideal solutions","link":"https:\/\/blog.skypilot.co\/slurm-vs-k8s\/","pubDate":"Wed, 30 Jul 2025 07:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/slurm-vs-k8s\/","description":"There are a lot of discussions happening in AI infrastructure right now. On one side, we have researchers who trained on Slurm in grad school, comfortable with sbatch train_model.sh and the predictability of academic HPC clusters. On the other side, we have platform engineers who’ve spent the last several years of their career mastering Kubernetes, building sophisticated cloud-native architectures for web-scale applications.\nThe problem? Modern AI workloads don’t fit cleanly into either world, and we’re watching both communities attempt increasingly creative solutions to bridge this gap."},{"title":"SkyPilot 0.10: Enterprise-Ready AI Infrastructure with SSO, Dashboard, Workspaces, and More","link":"https:\/\/blog.skypilot.co\/announcing-skypilot-0.10.0\/","pubDate":"Thu, 24 Jul 2025 06:00:00 -0600","guid":"https:\/\/blog.skypilot.co\/announcing-skypilot-0.10.0\/","description":"Announcing SkyPilot 0.10 - the largest release yet with enterprise-grade features."},{"title":"The Evolution of AI Job Orchestration. Part 2: The AI-Native Control Plane & Orchestration that Finally Works for ML","link":"https:\/\/blog.skypilot.co\/ai-job-orchestration-pt2-ai-control-plane\/","pubDate":"Wed, 16 Jul 2025 07:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/ai-job-orchestration-pt2-ai-control-plane\/","description":"This is Part 2 of our series on the evolution of AI Job Orchestration. In Part 1, we explored how Neoclouds are democratizing GPU access but leaving the “last mile” unsolved. Now we’ll discover how AI-native orchestration tools are bridging that gap.\nWe need AI-Native Control Plane for Any Infrastructure While Neoclouds, specialized GPU cloud providers, have solved the hardware accessibility problem by offering cost-effective, high-performance clusters with advanced networking like InfiniBand, they’ve left a critical gap: the orchestration layer."},{"title":"The Evolution of AI Job Orchestration. Part 1: Running AI jobs on GPU Neoclouds","link":"https:\/\/blog.skypilot.co\/ai-job-orchestration-pt1-gpu-neoclouds\/","pubDate":"Tue, 08 Jul 2025 07:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/ai-job-orchestration-pt1-gpu-neoclouds\/","description":"If you’re an infrastructure or MLOps engineer at a large company, you know the drill. The ML team comes to you with requirements that change weekly. They need GPUs yesterday, but the budget was set six months ago. They want to use the latest framework, but it breaks your carefully crafted Kubernetes deployments. They need to comply with data locality requirements while also optimizing for cost.\nSound familiar? You’re not alone, and there’s a better way."},{"title":"Managing Networks in the Chaotic Cloud and Kubernetes World","link":"https:\/\/blog.skypilot.co\/network-tier-on-multiple-clouds\/","pubDate":"Wed, 02 Jul 2025 00:10:19 -0700","guid":"https:\/\/blog.skypilot.co\/network-tier-on-multiple-clouds\/","description":"Configure high-performance networking on different cloud providers and managed infrastructure with unified SkyPilot's network tier abstraction"},{"title":"High-Performance Model Checkpointing on the Cloud","link":"https:\/\/blog.skypilot.co\/high-performance-checkpointing\/","pubDate":"Tue, 08 Apr 2025 00:00:00 +0000","guid":"https:\/\/blog.skypilot.co\/high-performance-checkpointing\/","description":"Techniques to speed up checkpointing by 9.6x and how to easily achieve them in SkyPilot"},{"title":"Large-Scale AI Batch Inference: 9x Faster Embedding Generation","link":"https:\/\/blog.skypilot.co\/large-scale-embedding\/","pubDate":"Thu, 20 Mar 2025 00:06:19 -0700","guid":"https:\/\/blog.skypilot.co\/large-scale-embedding\/","description":"How to accelerate distributed embedding generation? Use the "forgotten" regions."},{"title":"Introducing SkyPilot Client-Server Architecture","link":"https:\/\/blog.skypilot.co\/client-server\/","pubDate":"Mon, 10 Mar 2025 08:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/client-server\/","description":"Transforming SkyPilot into a scalable, multi-user platform."},{"title":"Abusing SQLite to Handle Concurrency","link":"https:\/\/blog.skypilot.co\/abusing-sqlite-to-handle-concurrency\/","pubDate":"Tue, 04 Mar 2025 08:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/abusing-sqlite-to-handle-concurrency\/","description":"SkyPilot uses the venerable SQLite for state management. SQLite can handle millions of QPS, and terabytes of data. However, our efforts to scale our Managed Jobs feature ran up against the one downfall of SQLite: many concurrent writers. Since SkyPilot typically runs as a CLI on your laptop, we wanted to stick with SQLite, so we decided to figure out how we can make it work. (Update: SkyPilot can now be deployed as a remote team server and supports PostgreSQL!"},{"title":"Using DeepSeek R1 for RAG: Do's and Don'ts","link":"https:\/\/blog.skypilot.co\/deepseek-rag\/","pubDate":"Wed, 26 Feb 2025 00:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/deepseek-rag\/","description":"DeepSeek R1 has shown great reasoning capability when it is firstly released. In this blog post, we detail our learnings in using DeepSeek R1 to build a Retrieval-Augmented Generation (RAG) system, tailored for legal documents.\nWe choose legal documents because legal professionals often face a daunting task: navigating libraries of cases, statutes, and informal legal commentary. Even the best-intentioned research can get bogged down in retrieving the right documents, let alone summarizing them accurately."},{"title":"Building Large-Scale Image Search using VectorDB & OpenAI CLIP: From 120 Hours to 1 Hour, From $$$ to $","link":"https:\/\/blog.skypilot.co\/large-scale-vector-database\/","pubDate":"Tue, 11 Feb 2025 00:06:19 -0700","guid":"https:\/\/blog.skypilot.co\/large-scale-vector-database\/","description":"SkyPilot enables image-to-image and text-to-image search from 120 Hours to 1 Hour and from $$$ to $"},{"title":"SkyPilot 0.7: 3x Faster Provisioning, Reserved Instances, Admin Features, New Hardware","link":"https:\/\/blog.skypilot.co\/announcing-skypilot-0.7\/","pubDate":"Thu, 14 Nov 2024 07:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/announcing-skypilot-0.7\/","description":"Announcing SkyPilot 0.7."},{"title":"Getting $1M cloud credits for AI startups \u2014 and using them wisely","link":"https:\/\/blog.skypilot.co\/million-dollar-cloud-credits-for-ai-teams\/","pubDate":"Fri, 01 Nov 2024 07:28:19 -0700","guid":"https:\/\/blog.skypilot.co\/million-dollar-cloud-credits-for-ai-teams\/","description":"For AI teams: How do you efficiently spend $1M+ cloud credits across 3+ clouds?"},{"title":"Can Multimodal LLMs Truly \"See\" Images? A Deep Dive with ASCII Art","link":"https:\/\/blog.skypilot.co\/can-multi-modal-llms-truely-see-images\/","pubDate":"Mon, 16 Sep 2024 00:00:00 +0000","guid":"https:\/\/blog.skypilot.co\/can-multi-modal-llms-truely-see-images\/","description":"With last week’s Pixtral release, multimodal large language models (LLMs) like OpenAI’s GPT-4o, Google’s Gemini Pro, and Pixtral are making significant strides. These models are not only able to generate text from images but are often touted for their ability to “see” images similarly to human perception. But how true is this claim?\nUpdate (2024-09-25): I added a new test for Vision Llama 3.2. Unfortunately, it failed again. Update (2024-09-18): I added a new test for Qwen2-VL, which still fails to generate the correct ASCII art."},{"title":"Finetune Llama 3.1 on Your Infra","link":"https:\/\/blog.skypilot.co\/finetune-llama-3_1-on-your-infra\/","pubDate":"Tue, 23 Jul 2024 00:00:00 +0000","guid":"https:\/\/blog.skypilot.co\/finetune-llama-3_1-on-your-infra\/","description":"Operational guide to finetune Llama 3.1, with everything packaged in a simple SkyPilot YAML."},{"title":"AI on Kubernetes Without the Pain","link":"https:\/\/blog.skypilot.co\/ai-on-kubernetes\/","pubDate":"Thu, 11 Jul 2024 07:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/ai-on-kubernetes\/","description":"Develop, Train and Serve AI on Kubernetes with SkyPilot."},{"title":"SkyPilot 0.6: Managed Jobs API, SkyServe on Kubernetes, Spot + On-demand mixing, Paperspace support","link":"https:\/\/blog.skypilot.co\/announcing-skypilot-0.6\/","pubDate":"Tue, 04 Jun 2024 07:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/announcing-skypilot-0.6\/","description":"Announcing SkyPilot 0.6."},{"title":"Introducing SkyServe: 50% Cheaper AI Serving on Any Cloud with High Availability","link":"https:\/\/blog.skypilot.co\/introducing-sky-serve\/","pubDate":"Tue, 20 Feb 2024 00:00:00 +0000","guid":"https:\/\/blog.skypilot.co\/introducing-sky-serve\/","description":"SkyServe: A simple, cost-efficient, multi-region\/cloud library for serving GenAI models."},{"title":"Scaling Mixtral LLM Serving with High GPU Availability and Cost Efficiency","link":"https:\/\/blog.skypilot.co\/scaling-mixtral-llm-serving-with-the-best-gpu-availability-and-cost-efficiency\/","pubDate":"Thu, 21 Dec 2023 00:00:00 +0000","guid":"https:\/\/blog.skypilot.co\/scaling-mixtral-llm-serving-with-the-best-gpu-availability-and-cost-efficiency\/","description":"A tutorial for serving Mixtral 8x7B model with SkyPilot and SkyServe."},{"title":"Scaling AI Robotics on the Cloud","link":"https:\/\/blog.skypilot.co\/covariant\/","pubDate":"Tue, 26 Sep 2023 08:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/covariant\/","description":"Covariant runs AI on the cloud using SkyPilot, delivering models 4x faster cost-effectively."},{"title":"Finetuning Llama 2 in your own cloud environment, privately","link":"https:\/\/blog.skypilot.co\/finetuning-llama2-operational-guide\/","pubDate":"Wed, 02 Aug 2023 00:00:00 +0000","guid":"https:\/\/blog.skypilot.co\/finetuning-llama2-operational-guide\/","description":"An operational guide on finetuning Llama 2, ready for commercial use."},{"title":"Serving LLM 24x Faster On the Cloud with vLLM and SkyPilot","link":"https:\/\/blog.skypilot.co\/serving-llm-24x-faster-on-the-cloud-with-vllm-and-skypilot\/","pubDate":"Thu, 29 Jun 2023 00:00:00 +0000","guid":"https:\/\/blog.skypilot.co\/serving-llm-24x-faster-on-the-cloud-with-vllm-and-skypilot\/","description":"SkyPilot makes the deployment and development of vLLM easy and fast on clouds."},{"title":"SkyPilot 0.3: LLM support and unprecedented GPU availability across more clouds","link":"https:\/\/blog.skypilot.co\/announcing-skypilot-0.3\/","pubDate":"Tue, 30 May 2023 07:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/announcing-skypilot-0.3\/","description":"Announcing SkyPilot 0.3: LLM support, new clouds, and enhanced production readiness."},{"title":"Analyzing the Whole Mouse Brain Atlas on the Cloud With SkyPilot [User Post]","link":"https:\/\/blog.skypilot.co\/analyzing-the-whole-mouse-brain-atlas-on-the-cloud-with-skypilot\/","pubDate":"Mon, 01 May 2023 09:00:00 -0700","guid":"https:\/\/blog.skypilot.co\/analyzing-the-whole-mouse-brain-atlas-on-the-cloud-with-skypilot\/","description":"Experience report from Salk Institute on how biologists use SkyPilot to conduct research on the cloud."},{"title":"Run LLaMA LLM chatbots on any cloud with one click","link":"https:\/\/blog.skypilot.co\/llama-llm-chatbots-on-any-cloud\/","pubDate":"Mon, 20 Mar 2023 00:00:00 +0000","guid":"https:\/\/blog.skypilot.co\/llama-llm-chatbots-on-any-cloud\/","description":"Want to host your own LLM Chatbot on any cloud of your choosing?"},{"title":"SkyPilot: ML and Data Science on any cloud with massive cost savings","link":"https:\/\/blog.skypilot.co\/introducing-skypilot\/","pubDate":"Wed, 16 Nov 2022 00:00:00 +0000","guid":"https:\/\/blog.skypilot.co\/introducing-skypilot\/","description":"Introducing SkyPilot."}]}}