Engineering velocity isn’t just about writing more code — it’s about reducing the cognitive load of finding the right context across GitHub, Jira, Slack, and countless docs. Glean’s pre-built engineering agents do exactly that, with purpose-built workflows for PR review, Jira resolution, onboarding, self‑evals, and more. Really excited about how this shifts AI from “cool demos” to actual infrastructure for how teams ship and fix software. https://lnkd.in/gp5EME6y

Engineering velocity isn’t just about writing more code — it’s about reducing the cognitive load of finding the right context across GitHub, Jira, Slack, and countless docs. Glean’s pre-built engineering agents do exactly that, with purpose-built workflows for PR review, Jira resolution, onboarding, self‑evals, and more. Really excited about how this shifts AI from “cool demos” to actual infrastructure for how teams ship and fix software. https://lnkd.in/eHq7rqBv

If your team is feeling the pain of slow reviews, messy launches, or painful performance/self evaluations, this is worth a look. @Glean’s latest Drop introduces engineering-focused agents that: 👉 summarize and standardize PR reviews 👉 propose resolution paths for tricky Jira tickets 👉 auto-generate launch docs and self‑evaluations from real work artifacts Instead of adding another tool to juggle, these agents sit on top of the systems you already use and give engineers back time to build. @Chenhao Yang from our engineering team demos a few of the agents: https://lnkd.in/gwasamff

Managing multiple agents can be a nightmare! Which is why I added a kanban to manage your sessions/agents on opencode. Create tasks → Plan when ready → Unblock agents → Start building → Review → Complete changes A nice and easy workflow. Checkout the video below

I realized something watching my team struggle with a feature launch last week: we were so focused on shipping fast that we forgot to ask why we were shipping at all. We had the roadmap. We had the sprints. We had the Jira tickets color-coded and prioritized. But somewhere between the kickoff and the demo, the original problem we were solving got buried under layers of requirements and assumptions. One engineer finally asked: "Do we actually know if customers want this?" Silence. Turns out, half the features we built were solving for internal conversations—not real user pain. We'd optimized for speed without validating direction. That week cost us. Not in money. In clarity. Now, before anything hits the calendar, we do a single thing differently: we start with the smallest possible question. "What's the one thing we need to learn?" Not the roadmap. Not the timeline. Not the design. The learning. It sounds simple. It's not. Because it means saying no. It means moving slower upfront to move faster later. It means admitting what you don't know. But every shipping cycle since? We've cut waste in half. Speed without direction is just noise.

Reposting per request - pleased to share. You're invited to join the upcoming workshop on Robotic Process Automation (RPA) - a powerful technology that is reshaping how modern organizations operate.   Robotic Process Automation is transforming how organizations operate by automating repetitive tasks, reducing errors, and freeing people to focus on creative, high‑value work. Companies across finance, healthcare, logistics, and tech are rapidly adopting automation tools, which means students with these skills stand out immediately.   Students who understand these tools gain a strong competitive edge in internships and job opportunities. This workshop will give you hands‑on experience, real examples, and practical skills that employers look for. 

Your standup asks "what are you working on today?" Wrong question. Should be "what's blocking you from shipping?" CodePulse tracks the invisible delays: PRs sitting in review limbo, code waiting for CI, deploys stuck in staging. The stuff that doesn't show up in Jira. Most teams know their velocity. Few know their actual cycle time. That gap is where productivity dies. We built CodePulse because engineering leaders kept asking "why are we slow?" but had no data to answer it. Now they do.

I’ve spent the better part of 15 years building automation and writing code to make work more efficient. Recently, I started using Code Puppy in a few of my workflows, and it’s quickly become one of the most impressive tools I’ve ever worked with. What stands out is how much heavy lifting it does on its own. Instead of spending time building out layers of error‑handling, Code Puppy actively works through problems, iterates, and finds solutions with very little intervention. Watching it reason through issues with its “cubs” is genuinely impressive — and a big shift from the way most automation tools operate. From what I’ve seen so far, there aren’t many repetitive tasks it can’t improve. The team behind it built something really special. If you have access to Code Puppy and haven’t taken the time to explore it, you’re leaving a lot of value on the table.

Reposting this from a different lens. I’ve never been a coder. For many years, automation and efficiency felt like something technical teams owned. Recently, I hit a workflow pain point and immediately thought, “We need to build an agent.” So I asked Code Puppy to build it. It told me it could — but there was a more efficient way to solve it in under five minutes. No coding required. And it gave me the steps. That was the unlock. What excites me isn’t just efficiency. It’s the shift in thinking. I’m more curious. I experiment more. I move faster from “Wouldn’t it be nice if…” to “Let me try this.” I now start with AI when I approach tasks and problems — not as a shortcut, but as a thought partner. And that has changed how I lead and how I solve. When tools can reason, iterate, and even challenge your first instinct, the barrier to entry drops. Innovation expands beyond developers to anyone willing to explore. For leaders driving transformation, that’s the opportunity: not replacing technical depth, but expanding who gets to participate in building solutions. Curious how others are leaning into AI-assisted workflows.

It’s incredible how Linear is eating Jira’s marketshare in a matter of months. All my dev engineering manager friends are switching. And the lesson here? If you fall asleep and don’t ship and iterate fast enough, you’re gonna get eaten. No matter your size or history.