Cloud spending hit $107 billion in Q3 2025, according to Synergy Research Group data. AWS around 29%, Azure 20%, and Google Cloud 13%.
Sure, they offer exceptional tech and service, right?
Well, let's take an orchestrator like Apache Airflow (no exception): it'll easily set you back €300+/month, whereas a comparable setup on self-hosted hardware (e.g., a VPS) can be €5/month. 2x I get it, 5x maybe, 10x that's greedy-but 100x more?
Self-hosting!
Each traverse_node is a separate step, created at runtime, with its own artifacts, retries, and lineage.
Avi Chawla (@_avichawla)
임베딩 스택이 모델을 바꿀 때마다 100% 재인덱싱을 강제한다는 문제 제기. 많은 팀이 이를 불가피하다고 여기며, 예를 들어 대형 임베딩 모델로 RAG 파이프라인을 구성해 프로덕션에 배포한 뒤 시간이 지나 트래픽·요구사항이 바뀌면 전체 재인덱싱 비용과 운영 부담이 크게 증가한다는 점을 지적하고 있음.
24 hours until the CfP for "LLMday NYC 2026 Q1" closes: https://papercall.io/cfps/6488/submissions/new
TechFollow (@TechFollowrazzi)
Micah Hill-Smith는 ArtificialAnlys의 공동창업자 겸 CEO로, 독립적인 AI 벤치마킹 플랫폼을 운영해 팀들이 특정 사용 사례에 맞는 최적의 모델과 API 제공자를 선택하도록 도와줍니다. 모델 평가·비교에 특화된 서비스라는 점이 강조됩니다.
LLM-generated EHR summaries can fail in the worst way: one confident hallucination that changes clinical meaning.
This guide shows claim-level evaluation, risk-weighted safety metrics, and production gates (generate → verify, conservative fallbacks, monitoring).
Read: https://codelabsacademy.com/en/blog/evaluating-llm-hallucinations-clinical-safety-ehr-summaries?source=mastodon
Dự án mã nguồn mở PardusClawer hỗ trợ data scientist, tích hợp Ollama & tự động tìm kiếm dữ liệu qua web. Cộng đồng Reddit giới thiệu công cụ đáng thử từ tác giả /u/jasonhon2013.
#DataScience #OpenSource #AI #Python #MLOps
#KhoaHocDuLieu #MaNguonMo #TruongHopSVM #HocMay #RedditVietNam
https://www.reddit.com/r/LocalLLaMA/comments/1qt05j3/openclaw_for_data_scientist/
Cloud is "stable", right?
2025 reminded us that even the biggest providers can fail spectacularly: outages at AWS, Cloudflare, Google and others affected *millions* globally and cascaded into major service disruptions. Data aggregated by Ookla paint a stark picture of how reliant we are on a handful of providers.
Do you think *small companies and devs* are taken seriously by these cloud giants?
Thinking about a new MLOX playground 🛠️
Recently, the new Raspberry Pi AI HAT+ 2 was announced. An accelerator add-on for the Raspi 5 that brings local generative AI capabilities to the Pi delivering ~40 TOPS of inferencing performance.
I'm tempted to order one now and see how far we can push local AI workflows on tiny hardware. Luckily, MLOX makes experimenting easy, I already tested it with raspi zero 2W which worked like a charm.
Next week: how ZenML approaches dynamic pipelines specifically.
🔮 Turning models into products: that’s when machine learning becomes real magic.
Francys Lanza shared with us how Bankaya built an enterprise-grade ML pipeline that automates deployment, retraining, and monitoring across risk, marketing, and more. This is how you turn experiments into enterprise capabilities.
🎥 Watch now! https://abracadata.space/sessions/2025/scaling-machine-learning-in-enterprises/
MLOps eng chán debug chi phí LLM tràn ngập đã xây ToolTokenTrail để kiểm soát. Tính năng nổi bật: Thống kê chi phí theo khách hàng, circuit-breaker ngắt request khi cost vượt ngưỡng,追踪 agent tiêu tốn chi phí, và route query sang mô hình rẻ hơn. Cần feedback về Outcome-Based Billing (chỉ tính phí khi thành công). #MLOps #LLM #AI #SAAS #TokenTrail #KinhDoanhTech #PhatTrienPhanMem
https://www.reddit.com/r/SaaS/comments/1qqdzpe/im_an_mlops_eng_tired_of_debugging_llm_cost/
In #MLOps for #HealthcareAI, “service is up” doesn’t mean the model is safe. Monitor drift, missingness spikes, calibration shifts, and API abuse without leaking PHI through logs.
Read the full guide → https://codelabsacademy.com/en/blog/healthcare-ml-monitoring-incident-response-models?source=mastodon
These aren't new problems. We solved (or at the very least surfaced and discovered) them in ML pipelines years ago. But as people productionise agents, all this infrastructure becomes critical again.
So my question isn't "are dynamic pipelines good or bad?" It's: when do you actually need them, and when are you just avoiding the work of understanding your domain?
More tomorrow on where dynamic genuinely makes sense.
MLOX is not:
• a one-click, no-code agentic-AI platform
• a YAML-generation framework
• a Kubernetes religion
• a VC-backed hockey-stick growth story
It's an attempt to make AI infrastructure boring enough that it doesn't ruin your week :clippy:
🧠 🆕 Just launched: Deploying Small Language Models (LFWS307)
A live, 3-day, instructor-led workshop with 10+ hours of hands-on labs.
Learn to deploy SLMs across laptops, servers, edge devices, and the browser—skills built for real-world AI infrastructure and ML platform engineering.
Enroll now: https://training.linuxfoundation.org/training/deploying-small-language-models-lfws307/
#AI #MLOps #Vietnam
Lỗi khi tự lưu trữ Kimi K2.5 trên 8x H200: Phản hồi ngẫu nhiên '(no content)', gọi công cụ thất bại trên vllm/sglang. Sử dụng opencode gây lỗi parser và hiển thị ký tự trúng <|...|> trong kết quả. Cần hỗ trợ khắc phục!
https://www.reddit.com/r/LocalLLaMA/comments/1qpce5w/self_hosted_kimi_k25_no_content_issue/
Why ai training data services Are the Real Foundation of Successful Machine Learning
Strong AI outcomes start with strong data. Inconsistent or poorly labeled datasets slow innovation and reduce ROI. AI training data services & reliable training data services help organizations build accurate, scalable models by ensuring data quality from the start. When data is right, AI decisions become faster, smarter, &more reliable.
https://www.hitechdigital.com/blog/data-preparation-for-machine-learning
One nice side effect of writing the CAIN paper:
It forced me to clearly articulate why MLOX exists and not just what it does.
That clarity is now feeding directly back into the code and docs.
If you enjoy working on open infrastructure and want to help shape this early, feel free to reach out or have a look at the repo 🙂 (mlox.org)
⚙️ 𝗟𝗲𝗮𝗿𝗻 𝗛𝗼𝘄 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝘃𝗲 𝗔𝗜 𝗥𝗲𝗮𝗹𝗹𝘆 𝗪𝗼𝗿𝗸𝘀 - #𝗡𝗲𝘄𝗕𝗮𝘁𝗰𝗵
🚀 2026 isn’t about learning AI basics. It’s about knowing how to use AI properly.
👨🏫 𝗧𝗿𝗮𝗶𝗻𝗲𝗿: Mr. Vivek
📅 𝗗𝗮𝘁𝗲: 05/02/2026 (Thursday)
⏰ 𝗧𝗶𝗺𝗲: 7:00 AM IST
🔗 𝗝𝗼𝗶𝗻 𝗹𝗶𝘃𝗲: https://teams.microsoft.com/meet/45827062316950?p=Cl5vezOETDeten5CJv
🆔 𝗠𝗲𝗲𝘁𝗶𝗻𝗴 𝗜𝗗: 458 270 623 16950 | 𝗣𝗮𝘀𝘀𝗰𝗼𝗱𝗲: Vz7sC7de
📞 𝗖𝗼𝗻𝘁𝗮𝗰𝘁: +91 7032290546
📲 𝗩𝗶𝘀𝗶𝘁: https://www.visualpath.in/generative-ai-course-online-training.html
#generativeai #genai #promptengineering #mlops #llmops #aiops #datascience #agenticai #cloudai #azureai #awsai #gcpai #aiml #onlinecourses
