DORA Webinar; 'Vibe Coding' & EMs; Effective Teams with GenAI; AI Engineer Vs Software Engineer; Model Context Protocol;
Issue #45 Bytes
đ± Dive into Learning-Rich Sundays with groCTO —ïž
The Hows & Whats of DORA ft. Mario, Sr. EM at Contentsquare
Slow cycle times, high work-in-progress, and unclear efficiency benchmarks hinder many engineering teams. The solution lies in effectively using DORA & other engineering metrics. In âThe Hows & Whats of DORAâ webinar powered by Typo, host Kovid Batra and metrics experts:
Mario Viktorov MechoulamâSenior Engineering Manager at Contentsquare
Kshitij MohanâCo-Founder & CEO of Typo
explore the science of engineering metrics, their evolution, and their impact on team performance, shipping velocity, and business outcomes. Here is a quick clip from the webinar and link to the full recording.
Article of the Week â
âThe LLMs will become better and better at complex coding tasks. Cursor and others have a growing goldmine of data. They know what doesnât work based on our frustrated prompts, and there will be great LLMs specifically for coding.â
How vibe coding will affect Engineering Managers
Anton Zaides helps us explore the new age of constraints that âvibe codingâ and similar new trends at the frontier of GenAI research is bringing on to established engineering teams.
DenisâEditorial Note: âVibe Codingâ was originally used by Andrej Karpathy as a casual demonstration of using voice-only prompts to have a chat bot control an IDE to produce a software prototype. Iâm aware that this has been vaguely expanded to also include AI-assisted coding and agentive usages in popular culture. Weâre presuming the broader term here.
The emerging reality isnât about replacement, but reorientation. With delivery speeds accelerating across the board, the traditional value of âhow fast can we shipâ is being replaced by âhow well can we choose what to build and own the outcomes.â Engineering managers are no longer just optimizing team throughput. Developers are being pushed into broader product ownership, while EMs must bridge strategy and execution.
The EMs who thrive will be those who can navigate complexity with clarity, support deeper product understanding across their teams, and build cultures that are resilient, autonomous, and aligned to business impact.
1. The Bottleneck Shifts Away From Engineering
As AI accelerates development, engineering stops being the bottleneck. Constraints will move upstream to product strategy, user research, and decision-making. EMs must now influence business outcomes, not just execute delivery. However, it has to be said that writing software wasnât ever the bottleneck in large-scale product engineering. GenAI can help great engineers broaden their scope, but can also accelerate the production of waste and issues in a less-optimal engineering culture.
2. Teams Will Own More Code, Systems, and Context
Faster development = more features = more surface area to maintain. Teams will need to manage growing tech scope with fewer people. AI helps with comprehension, but EMs must handle chaos, context-switching, and keeping things maintainable.
3. Expect More Bugs, Crashes, and Chaos
More code, less oversight, and AI-generated suggestions mean more surprises. Integration tests, automated checks, and good architecture matter more than ever. EMs are now the last line of defense.
4. Engineers Must Think Like Product People
AI handles more of the âhow.â That means engineers must own more of the âwhatâ and âwhy.â This demands stronger product sense, faster decisions, and less reliance on PMs for every small call.
5. Focus Time Will Shrink, Interaction Will Grow
Coding will be less about isolated flow and more about real-time collaboration. Pair programming, prompt-crafting, and co-creating with teammates (and AI) will become more common.
đ âPair vibingâ might be the new norm.
Other highlights đ
Leading Effective Engineering Teams in the Age of GenAI
AI is not here to replace engineers. It's here to reshape how we lead them.
Addy Osmani explores the intagibles and risks of GenAI chat bots in software engineering teams and their leadership. The real question for CTOs isn't âhow much faster can AI write code?â. It's âhow do we build better software, at scale, with AI as a collaborator?â
Key Concepts for Forward-Thinking Leaders
AI â Autopilot: Treat GenAI as a junior developer. Fast, helpful, but error-prone. Implement trust-but-verify as a team-wide mindset.
The 70% Problem: AI gets you 70% of the way fast. But the critical last 30% (edge cases, performance, domain nuance) still demands human judgment.
Upskilling is Leadershipâs Job Now: Effective leaders train their teams to use AI responsibly from prompt engineering to output validation and ensure fundamentals like debugging and system design donât atrophy.
Productivity Metrics Are Changing: It's no longer just velocity. Focus on code quality, maintainability, and developer growth. Track how AI saves time and how that time is reinvested (mentorship, innovation, tech debt cleanup).
AI Benefits Seniors More (For Now): Experienced devs extract the most from AI tools. Junior devs may appear productive, but lack understanding. Leaders must balance AI-accelerated output with learning depth.
Strategic Leadership Shifts
From Execution to Enablement: Your job is to set the why and what and coach the how, blending human creativity with AI capabilities.
Governance is Non-Negotiable: Implement clear policies for code review, data privacy, IP risk, and security when using AI tools. Human accountability stays central.
Ethics + Empathy Scale with Tech: As AI becomes a co-author of code, your human leadership becomes the differentiator.
What Great CTOs Do Differently
Champion âAI fluencyâ across the org
Protect engineering fundamentals and refactor your hiring, onboarding, and review processes to build understanding over output
Use AI to multiply developer impact, no one is getting replaced overnight
Build a culture of curiosity over fear to overcome the hype with sanity
AI Engineer vs. Software Engineer
AI engineers and software engineers differ significantly, from their focus on machine learning vs. application development to their development lifecycles. Did you know AI development is more experimental and iterative, requiring continuous model retraining, unlike traditional software's deterministic nature?
This blog dives into these key distinctions and addresses the question of whether AI will replace software engineers.
MCP (Model Context Protocol): Simply explained
Model Context Protocol is the bleeding edge of GenAI chat bots. Notably they allow GPTâs to access tools that can provide more context or raw data from a myriad of sources without having to copy/paste information into the context window manually, while also allowing the chat bot some semblance of control over what and how itâs querying new data or decisions.
Jordan Cutler simplifies MCPâs for us in this brief overview.
With MCP, AI coding assistants like Cursor gain contextual awareness, allowing them to perform actual development workflows like reading browser logs, fixing bugs, or creating tickets without manual input or API glue code. The impact? Faster debugging, tighter feedback loops, and AI-driven workflows that move beyond code generation into actual autonomous action.
For leaders, this signals a critical evolution: the rise of "agentic development environments" where AI tools not only write code but operate as full-stack collaborators. MCPâs standardized interface simplifies integration across tools, meaning teams can scale AI adoption without bespoke integrations or overhead.
Highlights
MCP turns LLMs into hands-on engineers by giving them access to real tools like Slack, Sentry, JIRA, and GitHub.
Say goodbye to brittle API glue code: MCP standardizes integrations so AI tools can work across systems without custom wrappers.
LLMs can now do, not just suggest: think debugging browser errors, scanning Slack alerts, and auto-generating PRs, all in one flow.
Cursor + MCP = AI that reads logs, tests fixes, and iterates, just like a developer would with no copy-pasting required.
Built on a familiar adapter pattern: MCP makes it dead simple to expose tools to AI via a consistent, declarative interface.
One
.jsonfile in Cursor, and youâre live: devs can wire up AI to their stack in minutes, not days.Early glimpse of agentic workflows: where AI doesnât just assist but drives tasks forward with autonomy.
MCP is the protocol powering AI-native development: expect this to become the foundation for how teams scale intelligent dev tooling.
Find Yourself đ»
Thatâs it for Today!
Whether youâre innovating on new projects, staying ahead of tech trends, or taking a strategic pause to recharge, may your day be as impactful and inspiring as your leadership.
See you next week(end), Ciao đ
Credits đ
Curators - Diligently curated by our community members Denis & Kovid
Featured Authors - Addy Osmani, Anton Zaides, Jordan Cuttler
Sponsors - This newsletter is sponsored by Typo AI - Ship reliable software faster.
1) Subscribe â If you arenât already, consider becoming a groCTO subscriber.
2) Share â Spread the word amongst fellow Engineering Leaders and CTOs! Your referral empowers & builds our groCTO community.