.png)
In our latest Tech Teams Today webinar, I sat down with three engineering leaders who’ve been living the reality of scaling tech teams in 2025:
- Kasey McCurdy – Head of Engineering at Gloo
- Pedro Tabio – Co-Founder & CTO at Vibrant
- Lalit Kundu – Co-Founder & CEO of Delty.ai
What unfolded wasn’t a theoretical conversation about AI or remote work—it was a tactical discussion about what’s working right now, what’s not, and what engineering leaders need to rethink if they want to scale effectively this year and beyond.
Here’s what stood out.
1. The Best Engineers Aren’t Just Coders
One of the strongest themes: writing code is NOT the most valuable skill on an engineering team. This has been true for some time now, but it's becoming increasingly important in an age where AI can do a lot of the grunt work.
Every panelist emphasized that modern engineers are increasingly being asked to think like architects. Problem-solving, system-level thinking, and making judgment calls based on tradeoffs are becoming core skills—even for mid-level and junior roles.
AI has made writing code easier and faster. But knowing what code to write—how to write is and why—is what separates high-performing engineers from average ones.
2. Don’t Hire Until You’re Drowning
All three leaders pushed back on the idea that growth equals headcount.
Their view? Hiring should come only after you’ve exhausted other options—like nearshoring, upskilling existing team members, reducing scope, or using contractors. The cost of hiring the wrong person, especially in a remote or early-stage environment, is simply too high.
For smaller teams, every hire changes the dynamic. For larger orgs, hiring too quickly often leads to slower delivery, not faster—because senior engineers end up spending their time reviewing and unblocking instead of building.
3. Curiosity, Adaptability, and Judgment > Credentials
Across the board, there was a strong bias toward hiring engineers who are self-directed, curious, and comfortable with ambiguity.
Rather than emphasizing degrees, years of experience, or traditional credentials, the panelists talked about evaluating how quickly someone can ramp up on new tools, navigate a fuzzy problem, or learn something on the fly. That skillset is more valuable—and more scalable—than raw knowledge.
4. Tools Are Evolving—But So Are the Evaluation Criteria
There was a lot of discussion about tooling—what’s actually helping and what’s just hype. Here’s a quick breakdown of where the panelists netted out:
Tools They’re Using and Liking:
- Cursor: Particularly powerful when customized with domain-specific rules. Used as a daily assistant for code generation, refactoring, and even debugging.
- CodeRabbit: A helpful AI reviewer that integrates cleanly into GitHub workflows. Some teams are using it alongside traditional review processes.
- Ingest: Valuable for AI agent workflow orchestration and queueing; also praised for its “agent kit” that supports more complex system logic.
- PostHog: Used by engineers (not just product teams) to own analytics, measure feature success, and stay aligned on outcomes.
- Graphite: For managing PRs and reviews more cleanly at scale.
- Linear: Praised not just for its design but as a model for how small, remote-first teams can ship world-class products.
Tools (and Patterns) They’re Moving Away From:
- Overreliance on LLMs without guardrails—especially when the organizational context isn’t baked in
- AI tools that feel like bolt-ons rather than integrated parts of the workflow
- Technical assessments that focus on rote knowledge or lead to code-style problems, rather than creative thinking and adaptability
- Old-school code review processes that bog down team velocity
There was a shared feeling that most teams are still figuring out how to balance AI augmentation with engineering intuition, and that the best tools are the ones that fit naturally into the way teams already work.
5. Process Isn’t Dead—It Just Needs to Be Lighter and Faster
Several panelists talked about how their teams have had to redesign processes to stay adaptable. Whether it was scaling down to regain focus, or rethinking sprint planning in a SOC 2-certified org, the common thread was intentionality.
You can’t scale well if you’re still operating with a playbook from 2019. That doesn’t mean abandoning structure—but it does mean building systems that allow for speed, experimentation, and fast feedback loops.
6. Remote-First Teams Require Deeper Trust and Higher Standards
Remote wasn’t just a background detail—it was central to how these leaders build culture and evaluate talent.
One takeaway: in a remote world, strong communication, presence, and initiative aren’t “nice to have”—they’re essential. Some concrete ways to improve communication: engineers are expected to turn their cameras on, jump into working sessions, and take responsibility for the features they ship—not just the code they write.
There was also a strong emphasis on embedding engineers into user feedback loops. Sitting in on customer calls, watching users struggle with your UI—these are the moments that make remote engineers feel connected to the problem they’re solving.
7. Leaders Need to Get in the Trenches
When change hits fast—and it will—engineering leaders can’t afford to hover. Everyone on the panel shared examples of getting deep into system design, writing code, or personally guiding their teams through shifts in architecture or team structure.
Leadership in 2025 isn’t about handing off responsibilities—it’s about removing blockers, setting direction, and modeling adaptability.
8. Fake Candidates Are Becoming a Real Problem
A concern that came up toward the end: fake candidates are showing up more often in remote hiring pipelines. With the rise of AI-powered resumes, identity obfuscation, and offshore proxy interviews, verifying who you’re hiring is becoming harder.
Leaders are investing more in identity verification and starting to redesign their vetting processes—not just to test skills, but to validate authenticity, initiative, and communication.
Final Thoughts
What struck me most about this conversation wasn’t the technology—it was how the engineering leader's mindset is evolving. The best teams in 2025 aren’t just the ones adopting AI tooling. They’re the ones who know how to hire for adaptability, how to scale without losing speed, and how to use new tools to amplify—not replace—engineering intuition.
If you’re leading an eng team right now—or planning to scale one this year—this conversation is worth your time.
Watch the full webinar recording: https://youtu.be/2CeAs9sWWII
Want to stay in touch with fellow engineering leaders? Join the T3 Exchange – our private community for engineering leaders: https://T3-Live.com/exchange
.webp){kind=small}
