Filter
Exclude
Time range
-
Near
As agentic engineering becomes more advanced, organizations are shifting their business models, including GitLab. Dive into this blog exploring the benefits of this shift, as well as what it means for customers. ➡️ spkaa.com/blog/revealing-the… #AgenticEngineering #DevOps #GitLab
5
The conversation at @NAFEMS Americas 2026 had a different tone than a year ago. Sessions on AI deployment outnumbered exploratory discussions, and the questions practitioners were asking were specific: how do we deploy agents responsibly, and what does our data foundation need to look like first? A few things stood out from the Rescale booth and the conference floor. 55% of attendees we surveyed are still manually passing results between simulation steps. Agentic workflows are beginning to address this, but trust is earned incrementally. On AI physics, the most consistent barrier practitioners cited was not the technology. It was data. Teams that delete simulation data to manage storage are cutting off the foundation that surrogate modeling requires. The full recap covers what practitioners are asking, what is working, and where engineering AI is heading over the next year. rescale.com/blog/nafems-amer… #NAFEMS #NAFEMSAmericas #AgenticEngineering #DigitalEngineering #EngineeringSimulation
1
32
Episodes 3 and 4 of our Agentic Engineering series with @PaoloRicciuti are out now, and they dig into the weeds about how to best keep an agent on track. Learn the right techniques for context management, and how test-driven development and git give you the guardrails to verify and iterate on everything the agent produces. Episode 3 👉 youtu.be/9vNj4YcGIw8 Episode 4 👉 youtu.be/eIHXU4lRKw0 #Svelte #SvelteKit #AgenticEngineering #FrontendDevelopment #ai
1
1
59
America's national labs have spent decades building some of the world's most powerful engineering simulation software. Today, we're making it accessible to U.S. industry at scale. Rescale is proud to announce the Agentic HPC Pipeline Initiative (AHPI), a proposed collaboration with three premier U.S. national laboratories: @BerkeleyLab, @ORNL, and @Livermore_Lab. For too long, these world-class DOE simulation codes have been out of reach for many manufacturers — requiring deep expertise to configure, run, and maintain on HPC infrastructure. AHPI is designed to close that gap. By pairing these codes with AI agents that automate complex workflows on the Rescale platform, American manufacturers can now explore design spaces, validate materials, and optimize manufacturing processes at unprecedented speed and scale. U.S. codes built on decades of national investment. Deployed for American industry. Read the press release: prnewswire.com/news-releases… #DigitalEngineering #AgenticEngineering #EngineeringSimulation #AdvancedManufacturing #HPC #NationalLabs

ALT Animated GIF of the first paragraph of Rescale’s recent press release appearing over a dark blue hexagon background, with the headlines “Press Release” and “Rescale and U.S. National Labs Form Initiative to Bring DOE Simulation Codes to American Industry”

1
5
1,386
Discover what it's really like to build production software, no hand-written code, just prompts and iteration. From launching two full-featured projects in two weeks, workflow tweaks, and honest challenges faced along the way. #AgenticEngineering #AI bit.ly/49hNOEq
2
73
"load-bearing" is the new AI-ism I'm already sick of. #ai #agenticengineering
12
#startabusiness 🇺🇸 🛸 ᵃᵍᵉⁿᵗⁱᶜ🛸ᵉⁿᵍⁱⁿᵉᵉʳⁱⁿᵍ 🛸 retweeted
A free 30-minute guide to mastering Claude Code: tinyurl.com/2s3pb6mr
1
1
29
Not for the weak. If you're a good software engineer, you'll catch them before they lay more 🐞🪺 #SoftwareEngineer #AgenticEngineering #AgenticAI #coding
5
🔥June is going to be HUGE for the Agentic AI Engineering community. I'm excited to be hosting two special events on two emerging disciplines one in London 🇬🇧 and one in San Francisco 🌉 that are rapidly shaping the future of Agentic AI and Agent Engineering! 🇬🇧 Agentic Engineering with @alibaba_cloud : London Agentic AI 📍 Tessl Alibaba Cloud 📅 June 25 🎙️Speakers from Alibaba Cloud Recombine Tomoro 👉 luma.com/hdygxmyx 🌉 Harness Engineering: State of the Art in Agent Harnesses: San Francisco 📍 AWS Builder Loft 📅 June 29 🎙️ Speakers from @arizeai @awscloud (AWS) @RapidFireAIHQ and @cocoindex_io 👉 luma.com/rtd0f6ka 🙏 Spaces are filling fast, don't wait until last moment. RSVP now #AgenticEngineering #HarnessEngineering #AIAgents #AIEngineering
2
3
242
Non-production: the work that never happened In April 2026 an engineer at Anthropic pointed Claude Code at a class of API errors nobody had owned for years. It shipped over 800 fixes and cut that error class by a factor of a thousand. The engineer's own estimate for the same work by hand: four years. The point is not speed. It is that this work was never going to be scheduled at all. It sat below the line where a human's time is worth spending. That line just moved, and almost no one has priced in what sits underneath it. For two years I have written about non-consumption: the customers a bloated SaaS stack over-serves and prices out, reachable now with something small enough to actually buy. This is the mirror image on the supply side. Call it non-production. Non-production is the work that, at normal time and normal cost, was never going to get built. Deferred bug fixes. The migration everyone postpones. Documentation. Technical debt that is real but never urgent enough to fund. None of it was worthless. It was simply below the hurdle rate. Be precise about what moved. The binding constraint was never labour hours. The engineer's words: solving other people's bugs is slow "because humans struggle to hold that much unfamiliar context in their head at once." That is a transaction cost in Ronald Coase's sense. agentic engineering collapses exactly that cost, which tells you which work gets unlocked first: high-context, low-glory, fragmented work. Migrations, cleanup, cross-cutting fixes. Here is the uncomfortable part, because it is the same force pointed the other way. When the marginal cost of a feature falls towards zero, the over-engineered product becomes the path of least resistance. Non-production is not only the cure for bloat. One level up, it is the cause of it. vibe coding a feature in an afternoon does not ask whether the feature should exist. So the logic inverts. When building gets cheap, not building gets valuable. The scarce resource is no longer production capacity. It is the discipline to decide what not to ship, and the judgement to tell a thousand-times improvement apart from a thousand things nobody asked for. The winner is not whoever produces most. It is whoever selects best. Full argument, with the Coase framing and the Monday-morning backlog test: drfloriansteiner.com/blog/no… Subscribe to the Weekly Agentic Engineering Digest at drfloriansteiner.com #AgenticEngineering #ClaudeCode #VibeCoding
3
1
129
Today’s Sunday What about your business #sunday #agenticengineering #chimirak #freelance
2
28
Your AI isn't bad at coding. Your harness is. 90% of devs are vibe coding: throw a vague prompt at the model, cross your fingers, hope it compiles. Great for a demo. Never for production. Here are 10 rules for going from tinkering to production-grade code : 1. The model is a brain, not an engineer. On its own it's blind. Your job is to build it a body : tools, CLI, hooks, skills, commands... 2. The model is stateless. It forgets everything between API calls. "Memory" is an illusion your infra recreates by re-injecting context every turn. 3. Kill the vibe coding. Don't let the model guess your architecture. Constrain its choices with strict schemas and rules. 4. Specialized agents beat one do-it-all agent. A generalist fails randomly. Spin up focused sub-agents (one for tests, one for docs), each with its own model and scoped permissions. 5. The orchestration triad. Command (orchestrator) → sub-agent (specialist) → skill (technical executor). The framework drives the workflow — not the model. 6. Watch out for the "dumb zone." Even with a 1M-token window, the model gets *lost in the middle*. Clear your session, prune your history. 7. The 200-line rule. Your source of truth (`CLAUDE.md`) stays short. Push everything else into local rules scoped per folder. 8. Progressive disclosure. Don't dump your whole codebase into context. The model reads only a skill's description and loads the details on demand. Protect your context window. 9. The universal loop: R-P-E-R-D.** Research → Plan → Execute → Review → Deliver. Five non-negotiable steps, whether you run BMAD, Spec Kit, or OpenSpec. 10. Version your harness. Your agents, system prompts, and skills are code. They live in your repo and keep the whole team aligned. 🛠️ Discipline turns a probabilistic model into a near-deterministic production tool. That's the whole point of agentic engineering: stop praying it works, start guaranteeing it. 👉 What are you orchestrating your agents with right now? Claude Code, a homegrown harness, something else? Drop it in the comments #AgenticEngineering #ClaudeCode #AI #DevTools #SoftwareEngineering
2
3
186