Your AI strategy should outlast your favorite model. A month ago, Guild’s engineering team was using Claude. A couple weeks later? Codex. Two weeks from now? It's anyone's guess. We don’t impose spending caps, and we don't have an opinion about how our engineers build. Our engineers gravitate where they see ROI. And that ROI math changes all the time. But here's the issue: your business processes, integrations, and operating requirements don’t. And if those investments become too tightly coupled to a single provider, the cost of switching can easily become a bottleneck to performance. For model selection to remain flexible, your infrastructure needs to remain independent. Read more from Akshay Punhani on this topic here: guild.ai/blog/engineering/wh…

Who's accountable for what your agents can do? Of the 362 U.S. IT decision-makers who responded to our recent Agent Management Report, the majority still couldn't agree on who actually owned their agents. Engineering/IT scored highest at 28.2%, 3.9% reported no single ownership standard -- and in a separate question, about half of all respondents admitted their organizations didn't have a formal policy governing agent use at all. Step 1 is proving that agents can do the work. Step 2 is proving they can do the work with the same visibility, accountability, and operational rigor of any other production system. Have you read the report? What stuck out to YOU? Let us know in the comments! guild.ai/blog/agent-governan…

New resource: download the Build or Buy Guide for Software Factories to explore— - key principles behind the build or buy debate - common pitfalls in both building and buying - and a total-cost calculator to measure potential costs against future ROI guild.ai/software-factory-bu…

AI spend is growing. Most leaders still don't know where. We'll be on the ground at @WeAreDevs 2026 (in San Jose, from Sept 23–25). If you want to learn how to control your AI spend, stop by the booth, check out talks from co-founders @jevering and Chris Waterson, OR book time with us here → guild.ai/we-are-developers-2…

We had a blast last night chatting all things agent operations at our first ever Guild Table Dinner. Shoutout to all the smart minds who came out to share in the conversation (and the hors d’ouvres). If you’re at #splunkconf26 looking for an amuse bouche from the Guild team, grab some time at the link below and we’ll give you a bite-sized demo of what we’ve been cooking up this summer in the Guild forge. guild.ai/conf-2026

@Splunk's .conf26 is here! Join us tomorrow (Tuesday) 2:15 PM - 3:00 PM MDT in Bluebird Ballroom 3C to watch Spencer Wanlass and Bhanu Prakash Kovuri leverage Guild + Splunk to guide a real agent from first-commit to continuously self-improving—live from the breakout stage. Deployed on Guild. Evaluated in Splunk. Wanna find out how you can create your own self-improving agents with Guild? Schedule some time to chat here: lnkd.in/eePFX7Bh #splunkconf26

Rethinking your AI infrastructure? Wondering what Guild.ai is all about? Guild CEO @jevering is going LIVE @ 8:30 PT w/ hosts @RonfortMartin and @MichaelWaitze on @AGTPinsights!

ICYMI check out @jevering , @ChrisWaterson , and the rest of the Guild team on @insidestartups with @willsclips_ for an inside look at the platform, the office, and the many bespectacled builders crafting the next generation of AI infrastructure. youtube.com/watch?v=IyaPJtR3…

34% of merged PRs at Guild are now written by Software Factory, autonomously. 56% of all code fixes come from it. 91% of merged Factory PRs need no additional engineer commits. Software Factory is changing how we build at Guild. It’s not another coding agent. It’s a system of specialized agents that continuously finds work, plans it, implements it, tests it, reviews it, and hands a finished PR back to an engineer. And it runs all the time. A few more numbers from our own codebase: - 70% of fixes for auto-detected issues come from the Factory - 81% of log-triage fixes are Factory-written - 80% of code changes to the Guild CLI are written by the Factory - 75% of Factory PRs are accepted and merged - 25% of alerts have a fix PR up within 90 minutes - 97% reduction in time-to-fix for some workflows - About $12 in model spend per merged Factory PR The important part is what those numbers add up to. The Factory is taking entire categories of engineering work off the team: production bugs, flaky tests, noisy logs, small fixes, and maintenance work that would otherwise sit in a backlog or interrupt an engineer. Humans still merge every PR. That’s deliberate. Agents do the work. People own the outcome. Coding agents were the first step. The next step is engineering systems that can continuously find, route, execute, review, and maintain work. We built ours for ourselves first. Now Software Factory is available for other teams to run on Guild. Try Software Factory → app.guild.ai/factory

It’s 2026. Do you know where your AI budget is? Last week we launched our new AI Spend Explorer — completely for free. Connect your platform provider and in <2 minutes, we’ll show you where every dollar went (across providers, models, users, teams, and agents) with a projection for the rest of the month. Files are read-only. There’s no credit card required. And setup is a single copy-paste. But why build this? And more importantly, why make it free? AI spend is growing fast, but most companies still can’t tell you where that money is going—or more importantly, if they could have paid less. Which agents are creating value? How much do they cost? And how do you make them dramatically more efficient without making them... worse? First, we’ll show you the bill. Then we'll can help you act on it. Visit spend.guild.ai to try it for yourself.

“The question I’m asking is: are we measuring work or value? An agent doing more work doesn’t necessarily mean the business is getting more value. Most organizations can tell you what they’re spending. Very few can tell you what they’re getting in return. That’s the gap I think the industry needs to close.” @jevering ICYMI, check out Guild CEO James Everyingham's tokenmaxxing convo with @LinearB_Inc's DevInterrupted, alongside @LaunchDarkly's @cetezadi and @TryArcade's @AlekzSalazar . devinterrupted.substack.com/…

Are you headed to @aiDotEngineer Worlds Fair? 6K+ attendees. 300 speakers. 100 AI labs on the expo floor. And one problem that's going to be top of mind everyone... Agent governance. If you're an enterprise leader asking how to govern agents effectively at scale, Guild's own Ben Oren and Corbett Waddingham will be onsite all week, and they would love to connect! DM Ben or Corbett on LinkedIn to grab a time 🗓️ Ben: linkedin.com/in/bforen/ Corbett: linkedin.com/in/waddingham/

Have you seen the new Guild Insights Dashboard? It's the first single view for your agents to understand at a glance how they consume resources, drive usage, and incur costs in real time. Check out this quick demo to find out how teams on Guild can: - See spend in dollars—to understand real financial impact, not just the token counts. - Trace costs to the source—to visualize spend by workspace, provider, model, and user. - Track spending by period — to measure and compare costs across a 7/30/90-day window. - And identify waste — to optimize token costs and reduce spend over time. Want to learn more? Check out our latest engineering blog: guild.ai/blog/product/insigh…

There's a storm coming. As agent sprawl takes hold in the enterprise, it’s creating a perfect storm of increased complexity and decreased visibility that's poised to turn the current agent governance gap into an agent governance crisis. - Runaway token costs - Agent lifecycle management - Unmanaged security risks - And lack of organizational accountability In our latest article, we discuss four of the biggest agent governance challenges facing enterprise teams today and why better infrastructure is the antidote. guild.ai/blog/ai-insights/th…

State of Agent Governance—June 2026 According to @Gartner_inc , the average Fortune 500 will deploy more than 150K AI agents by 2028. Only 13% believe they have the right agent governance to support it (@McKinsey ). There's no doubt, agent usage is growing fast; but the systems to govern it responsibly still aren't keeping pace. As agent sprawl takes hold in the enterprise, it’s creating a perfect storm of increased complexity and decreased visibility--and left unchecked, it's poised to turn a little agent governance gap into a big agent governance crisis. In our State of Agent Governance report, we review four of the biggest agent governance challenges facing enterprise teams right now and how the shadow of agent sprawl is exacerbating their impact: - Runaway token costs - Agent lifecycle management - Unmanaged security risks - Lack of organizational accoutability Check out our analysis of agent governance and why we believe better infrastructure will be the key to better outcomes in 2026 . Full report here: guild.ai/knowledge/ai-insigh…

Controlling agents is top of mind for the enterprise and @GuildAI CoFounder @jevering is sharing his perspective after 40 years in the industry. Check him out on the @mlopscommunity podcast with @Dpbrinkm to learn what AI investments are revealing about security and why access control is essential infrastructure for productizing agents safely at speed. Full episode available here: Spotify—open.spotify.com/episode/2MX… MLOps Community— home.mlops.community/home/vi…

How you build your agents has a lot to say about how you'll govern them as well. In his latest, Guild engineer @text2goat will walk you through how he builds an LLM agent on Guild's control plane.

@Microsoft just turned off Claude Code because the usage costs were too high. @ServiceNow blew through their entire 2026 token budget before Memorial Day. @Uber too. @Gartner_inc now predicts that over 40% of agentic AI projects will be canceled by the end of 2027, not because the models got worse, but because the economics and the risk profile never closed. This isn’t a value problem. It’s not even a pricing problem. It’s a discipline problem. You can’t control spend until you know how much you’re spending. And you can’t know what you’re spending if you don’t know what agents you’re accountable for in the first place. - Audit trails.  - Tracked permissions. - Clear and enforceable ownership standards.  - And the insights to respond quickly when your agents fail. This is the neutral governance layer that agents have been operating without since GPT 3. We decided it’s time to build it.

In a world full of surprises, sometimes you just need a little determinism in your life. In this quick demo, @text2goat will show you how to create deterministic type-script agents with Guild—and satisfy your maniacal need for control.