BairesDev

Most teams are using AI to do the same work faster. That's the smallest version of what it's for. The bigger shift is treating it as an enabler. An accelerant gets the same work done faster. An enabler changes what's worth working on. The more interesting question is what becomes possible that wasn't before: new problems worth solving, new skills, new roles that didn't exist a year ago. Most organizations are still on the first version. The second is where it gets interesting. This is one piece of a longer conversation with David Von Dollen, our Head of AI, on where adoption is actually heading. He breaks down the gap between the two in the clip. ▶️

Starting soon: "When Seeing Isn't Believing," a live conversation on securing identity and trust when AI agents can act on their own and deepfakes can convincingly imitate anyone. Join us at 1 PM ET for a candid look at what's breaking in production, and what security leaders are actually doing about it. On the panel: •Jason Conley, Chief Technology Officer at VIP Medical Group •Amanda H., Founder & Fractional CISO/CIO at Fiddlers •Patrick Doliny, Chief Information Security Officer at dōlogiic •Travis Rosiek, Public Sector CTO at Rubrik ➡️ Last chance to register: baires.dev/webinars-csol

The World Cup is over, but football is still on everywhere, and no two countries play it the same way. Software is no different. We ran our Q2 2026 Dev Barometer across 1,569 developers in 77 countries and built a starting XI: 11 countries, each defined by what its developers do best or fear most as AI reshapes the work. No single winner, just 11 different ways of playing the same moment. The common story is that AI is changing software work everywhere. That's true, and it hides how differently the change is landing. Brazil holds the line on fundamentals: only 12% of its developers rank AI tool fluency as the top hiring skill, the lowest in the group. Costa Rica bets the other way, with 15% crediting AI tools as the single biggest factor in landing their first developer job, 2 to 7 times higher than anywhere else. The wider numbers hold the same tension. 45% of developers globally still name problem-solving as the top hiring skill, more than double the 18% who pick AI fluency. And on what AI does to the career path, 57% of junior developers globally see it restructuring the old junior-to-senior climb, while 24% see it compressing it. Across the Latin American countries in the lineup, one pattern runs steadier: confidence outweighs worry. It's the most balanced read of a moment everyone is still figuring out. So, for anyone building a team right now: which position are you hiring for? Fundamentals, AI fluency, or the developers who keep shipping when the rules move.

Most boards are more afraid of a rogue AI agent going off-script than a well-executed deepfake call from "the CEO." That fear might be backwards. That's one of the questions we're putting to our panel on July 29 at 1 PM ET, alongside: •The convergence threat: how synthetic identity and agentic AI compound risk •Executive impersonation and voice cloning: real attack patterns, real responses •AI agent permissions and identity governance •Runtime monitoring and auditability •Rebuilding verification infrastructure for a world where trust can't be assumed Bringing their perspective to the table: Jason Conley (CTO, VIP Medical Group), Amanda H. (Founder & Fractional CISO/CIO, Fiddlers), Patrick Doliny (CISO, dōlogiic), and Travis Rosiek (Public Sector CTO, Rubrik). ➡️ Save your seat: baires.dev/webinars-csol

For years, cybersecurity meant protecting systems from attackers who looked like attackers. That's changing. AI agents can now take action with little human oversight. Deepfakes and voice cloning can convincingly imitate the people organizations trust most. The assumption behind most security models no longer holds: that a person is who they claim to be, and an AI is doing what it was told. On July 29 at 1 PM ET, join us for "When Seeing Isn't Believing," a live panel on rebuilding trust and governance in an AI-driven enterprise. On the panel: •Jason Conley, Chief Technology Officer at VIP Medical Group •Amanda H., Founder & Fractional CISO/CIO at Fiddlers •Patrick Doliny, Chief Information Security Officer at dōlogiic •Travis Rosiek, Public Sector CTO at Rubrik Register here: baires.dev/webinars-csol

Past a certain size, adding engineers to a team can slow down delivery. The communication math is unforgiving: a 5-person team has 10 communication channels, a 9-person team has 36, and a 15-person team has over 100. Each new channel is another place for delay and misalignment to accumulate. Past that point, plan for closer to 0.7 to 0.8x return on each engineer you add. In complex systems, that is often a strong outcome. Gilberto de Denghy, a project manager with 25+ years of leading delivery, notes that most scaling conversations start with open roles and budget. Few start with ownership boundaries, deployment pipelines, or review queues, which is usually where delivery is stuck. His piece lays out engineering team scaling strategies that run in sequence: diagnose the real bottleneck first, design small pods with clear ownership, and bring in nearshore engineers as true owners rather than a task queue. When delivery slows on your team, is the first move to add engineers, or to look at where ownership and review queues are actually stuck? ➡️ Full analysis here: https://lnkd.in/dBdmhF_N

Software outsourcing is growing straight through the tech hiring slowdown. Worldwide IT spending is set to rise 10.8% in 2026, while CFOs' headcount growth expectations have fallen from 6% to 2% (Gartner). Budgets are expanding faster than teams can hire, and that gap is landing on external engineering capacity. A second force is reshaping where that capacity sits. As of September 2025, a $100,000 supplemental fee on new H-1B petitions for workers outside the U.S. has made the onsite-heavy staffing model that defined outsourcing for two decades economically hard to justify. Robert Leibholz, our SVP of Business Development, maps the software outsourcing trends driving the shift: record contract activity in the Americas ($65.9B in 2025, up 25% YoY, per the ISG Index), and a regional split that's beginning to separate delivery models by what each one is actually good at. His read on where the work goes next: distributed-first delivery by default, with the regional choice increasingly turning on time-zone overlap and collaboration speed, not only the lowest labor cost. When you're deciding where delivery capacity sits, is cost still the deciding factor, or has time-zone overlap moved up the list? Full analysis here: https://lnkd.in/daYWY_Q6

Governing autonomous AI at scale is an infrastructure problem before it's a technology one. Most retail organizations aren't built for it yet. The models are commodities. What separates production from a pilot that never ships is everything around them: whether you can trace how an agent reached a decision, hold its actions accountable, and connect it to the systems that came before it. On July 8 at 1 PM ET, Colin Bodell (Liminal Velocity), William Groves (Bain & Company), Krishna Vedula (WellnessLiving), and Rajesh Rudraradhya (Lytx) join BairesDev to work through what closing that gap actually takes: what has to be in place before autonomous agents can run in a live retail environment, not just a demo. Register: baires.dev/webinars-csol

Most retail AI projects don't stall at the pilot. They stall at the integration layer. Legacy systems, fragmented data, and governance gaps aren't edge cases. They're the rule, and they're what separates a successful demo from production-ready AI. On July 8 at 1 PM ET, technology leaders are joining BairesDev for a candid operational conversation on what it actually takes to get AI out of the pilot phase and into production: - Colin Bodell - William Groves - Krishna Vedula - Rajesh Rudraradhya Reserve your spot: baires.dev/webinars-csol

For over 15 years, BairesDev has worked to connect the right engineering talent to the companies that need them. That same belief extends to the leaders running those teams. Today, E³ (Engineering Executive Exchange) launches with a founding cohort of 40 CTOs, CDOs, and VPs of Engineering from companies like Cisco, Adobe, and ServiceNow. E³ is an invite-only peer community built around recurring small-group conversations. Not a conference, not a platform. A room where senior leaders can think through the decisions that don't have an obvious peer inside their own company. The operating environment that those leaders are navigating is well-documented. BairesDev's AI Execution Gap report found that 79% of U.S. senior tech leaders feel pressure to overstate AI progress to satisfy stakeholders. A survey of senior technology leaders found that 97% have felt isolated in the role at some point. These aren't isolated findings. They reflect a structural condition at the top of engineering organizations: high-stakes decisions, compressed timelines, and very few people who share the same vantage point. BairesDev organized the community and brought together the founding cohort. The conversations belong to the members. Learn more about E³ at joinecubed.com