The Tech Trek

Rick Doten, cybersecurity startup advisor and AI researcher, joins the show to unpack how AI-assisted development is reshaping software—and what it means for security. From startups rushing to ship faster code to the unseen risks of “vibe coding,” Rick explains how engineering teams can balance innovation with secure, resilient design.

If your dev team is using AI tools to boost velocity, this conversation might change how you think about your SDLC, code review, and even your threat model.

Key Takeaways

• AI-assisted coding speeds up output but can multiply security risks if context isn’t baked in.

• Startups often trade speed for security early on—and that can be expensive to unwind later.

• Traditional fundamentals like OWASP and BSIMM still apply, even as architectures evolve with agents and MCP.

• AI creates a widening gap between companies that can secure their models and those that can’t.

• “Vibe coding”—non-devs using AI to build—introduces a new wave of shadow code leaders must prepare for.

Timestamped Highlights

[02:09] The real range of how startups are using AI-assisted tools—and why security is often an afterthought.

[05:12] Why AI-generated code is not just another form of third-party code.

[09:40] The hidden risk: code volume grows faster than your ability to secure it.

[15:51] How AI is widening the gap between resource-rich enterprises and everyone else.

[18:25] The new fragility of systems—where architecture and resilience start to break.

[22:07] Rethinking SDLC: integrating AI tools without losing security fundamentals.

[25:29] “Vibe coding” and what happens when non-engineers start shipping code.

Memorable Insight

“AI isn’t lazy like humans—it doesn’t just fix one thing. It rewrites everything. That’s why every line has to be re-scrutinized.”

Pro Tips

If your startup doesn’t have a dedicated security function yet, start with the basics: integrate OWASP checks into your CI/CD, use non-human accounts correctly, and automate code review gates early. Don’t wait until production to harden your systems.

Call to Action

If this episode sparked ideas for your dev or security team, share it with someone who’s experimenting with AI-assisted tools. Follow The Tech Trek for more conversations at the intersection of engineering, AI, and leadership.

What happens when a telehealth CTO takes AI beyond code generation and into the heart of the software development lifecycle?

Matt Buckleman, Co-founder and CTO of Hone Health, joins to share how his team uses AI not just to accelerate development, but to rethink workflows—from documentation and traceability to sentiment analysis across teams. This episode dives deep into how he’s blending engineering fundamentals with modern AI agents to create a smarter, more adaptive SDLC.

Key Takeaways

• Why AI’s biggest near-term value isn’t in code generation—it’s in improving process and communication.

• How Hone Health evolved its SDLC from three engineers on Slack to a 30+ person organization using agent-based automation.

• The hidden advantage of consistent naming conventions and traceability when applying AI to production systems.

• How AI can automate the “soft” but essential parts of software delivery, like documentation, requirements gathering, and developer sentiment tracking.

• What it takes to create feedback loops that make AI genuinely useful inside technical workflows.

Timestamped Highlights

[02:09] Flexible, anti-dogmatic SDLC: why strict process frameworks can slow learning.

[09:00] When more engineers doesn’t equal more output—the hidden cost of coordination.

[13:00] AI for experts vs. juniors: why prompting mirrors domain mastery.

[18:38] Offloading the unglamorous work: how LLMs now handle code comments, documentation, and swagger generation.

[23:50] Shared ownership and experimentation: how Hone’s engineering team pilots new AI tools.

[28:40] Turning meeting transcripts into smarter requirements: how agents refine specs automatically.

[32:00] Using sentiment analysis to spot risk and burnout across engineering projects.

Memorable Line

“LLMs are great at patterns in text—and that makes them better than people at understanding what’s really happening inside your workflow.”

Call to Action

If you enjoyed this conversation, follow The Tech Trek on Spotify or Apple Podcasts for more real-world discussions at the intersection of AI, engineering, and leadership. Share this episode with a teammate rethinking their own SDLC.

Yosi Dediashvili-Drossos, Co-Founder and CTO of City Hive, joins Amir to unpack how a hyper-focused approach helped transform a niche idea into the dominant e-commerce platform for the liquor industry. From bootstrapping into a complex, highly regulated space to giving small brands a voice, Yosi shares how City Hive built the connective tissue across the entire alcohol supply chain—bridging brands, distributors, and local retailers through data, trust, and mission-driven execution.

Key Takeaways

• Why narrowing your focus often creates more growth than going broad

• How City Hive turned regulatory complexity into a competitive advantage

• The power of connecting all layers of an industry—brands, distributors, and retailers—through one platform

• Why small, single-SKU brands now have a real chance to compete

• What founders need to know before tackling a regulated industry

Timestamped Highlights

00:36 – The origin story: building an e-commerce engine for liquor stores

04:00 – When niche focus becomes a gateway to full-scale growth

06:49 – Why the liquor supply chain is one of the most fragmented in the U.S.

10:22 – The uphill battle for small brands trying to reach consumers

12:16 – Empowering micro-brands through digital visibility and data

16:42 – How narrowing your scope can actually open new opportunities

19:48 – Lessons from scaling in a regulated market

22:49 – Yosi’s advice for founders navigating complex industries

Standout Moment

“You can’t solve everything at once. Focus on the next real problem that’s in front of you—if you do that well, you’ll eventually build something that can solve the bigger picture.”

Pro Tips

For founders entering regulated markets: Don’t start by trying to fix the system. Start by understanding one piece of it deeply enough that you can actually move it forward.

Call to Action

If you enjoyed this episode, follow The Tech Trek for more conversations with founders building technology that powers real-world industries. Share this episode with someone tackling a complex market—there’s a lot they’ll take away.

What happens when a 17-year Google veteran starts over with a 10-person AI startup? David Petrou, founder and CEO of Continua AI, joins Amir to unpack what it really takes to go from Big Tech stability to startup chaos. They dive into what to keep, what to unlearn, and how to build a high-performing team when everyone has to wear ten hats.

From career ladders to “vibe coding,” David shares a candid look at the tradeoffs, mindset shifts, and hard lessons behind scaling something new in AI.

Key Takeaways

• Career ladders are a luxury—startups win by hiring for adaptability and shared ownership, not rigid progression.

• Moving from Big Tech to startup means trading resources for speed—and rediscovering why building things is fun again.

• Productivity at small teams thrives on decisive action and ruthless prioritization, not endless debate.

• AI is transforming software development—but human experience still defines whether the tools actually deliver.

• The best retention strategy in a startup: keep the work interesting and the problems worth solving.

Timestamped Highlights

[00:48] How Continua AI brings “social AI” into group chats

[05:35] Why hiring for collaboration beats hiring for raw talent

[08:51] The real gap between Big Tech engineers and startup engineers

[11:19] What David had to unlearn after 17 years at Google

[18:58] How limited resources force sharper technical decision-making

[22:32] Productivity at early-stage startups—making faster decisions and moving forward

[26:41] “Vibe coding,” AI-assisted development, and why experienced engineers adapt faster

Memorable Moment

“It’s much better to be a few degrees off from optimal and moving fast than stuck in indecision for two weeks.” — David Petrou

Pro Tips

When hiring for an early-stage startup, focus less on titles or ladders and more on whether the person thrives without structure. The ability to figure things out independently is the best predictor of success.

Call to Action

If this episode gave you a fresh take on startup leadership, share it with someone thinking about making the leap from Big Tech to founder life. Follow The Tech Trek for weekly insights from leaders shaping the future of tech and AI.

When you step into a new leadership role, do you prefer to build a team from the ground up—or inherit one that already exists?

Ashwin Baskaran, VP of Engineering at Mercury, joins the show to unpack what really changes between these two scenarios—and what stays the same. From managing team dynamics to molding culture and earning trust in the first 90 days, Ashwin shares practical frameworks every engineering leader can apply.

Key Takeaways

• Building and inheriting share more similarities than most leaders realize—the principles of empathy, awareness, and low ego are universal.

• When inheriting a team, awareness is your first superpower. Learn the organization before making moves.

• Building from scratch gives freedom, but also more ways to make mistakes if you over-index on hiring people who think like you.

• The best leaders telegraph intent early and seek alignment through action, not reassurance.

• Feedback should be about context and priorities, not personal validation—it builds credibility and trust faster.

Timestamped Highlights

00:45 — The hidden overlap between building and inheriting a team

03:25 — Why self-awareness and low ego are critical when replacing a leader

06:51 — How “building” can lead to blind spots if you hire for similarity

11:38 — Finding alignment between company values and your leadership style

15:25 — How to read the room and earn feedback in your first 90 days

21:47 — What to look for when interviewing for a role where you’ll inherit a team

A Line That Stuck

“You want to find a problem that the team and company care about—and solve it in a way that feels aligned with their values.”

Call to Action

If this conversation helps you think differently about leadership transitions, share it with someone who’s stepping into a new role. Subscribe to The Tech Trek for more conversations that bridge technical leadership with real-world growth.

Jarah Euston, Co-Founder and CEO of WorkWhile, joins the show to share how she’s building a worker-first labor marketplace that puts money back into the pockets of frontline employees. Drawing from her own early experience in hourly jobs, Jarah explains why this massive yet underserved workforce deserves better tools, more respect, and faster access to earnings. We dive into automation, AI, re-skilling, and why the future of work isn’t just about robots replacing people but about using technology to unlock opportunity for 80 million Americans.

Key Takeaways

• Why hourly workers are overlooked in tech innovation and what WorkWhile is doing to change that

• How automation can cut overhead and actually raise wages instead of lowering them

• Why entry-level white-collar roles may be more at risk from AI than frontline jobs

• The importance of re-skilling and flexible training for workers who can’t stop earning to learn

• How instant pay and eliminating predatory fees can transform financial stability for families

Timestamped Highlights

01:26 — Jarah’s early jobs in retail and fast food and how they shaped her perspective

06:56 — Why frontline workers are less likely to be displaced by AI than software engineers

11:23 — Building against the grain: focusing on people instead of replacement tech

13:31 — Why robotics companies still hire frontline workers alongside automation

17:47 — Launching the American Labor Utilization Rate to track real work happening now

21:44 — Three pillars of WorkWhile’s mission: earning, upskilling, and financial access

25:17 — How word of mouth drives organic growth among workers and families

Memorable Line

“Even the companies building the future of automation still need people—and they’ve been our customers since day one.”

Call to Action

If this conversation opened your eyes to the future of frontline work, share it with someone who should hear it. Subscribe to the show for more conversations with founders and leaders reshaping technology and work.

Tom Drummond, Managing Partner at Heavybit, joins the show to break down what it takes to build and scale AI “picks and shovels” companies for the enterprise. We dive into the realities of selling into one of the hardest markets to reach, why differentiation matters more than ever, and how startups can wedge their way into massive opportunities despite fierce competition.

Key Takeaways

• Enterprise attention is more competitive than ever—breaking through requires clarity and category creation.

• Cold email and traditional outbound are saturated—startups must iterate quickly on channels and messaging.

• Landing enterprise deals often starts with developers and end users, not CIOs—grassroots adoption is powerful.

• Narrow wedges matter—solve one painful, high-value problem better than anyone else, then expand.

• Timing the industry cycle is critical—knowing when markets fragment and when they consolidate can define outcomes.

Timestamped Highlights

02:03 — Why enterprise attention has never been harder to win

04:55 — Differentiation in a sea of lookalike AI infrastructure startups

07:34 — Cold email vs content, billboards, and unconventional channels

08:35 — The Pareto rule of enterprise revenue and why developer adoption is key

11:47 — Competing with big tech incumbents: the power of the narrow wedge

21:03 — Where the market is headed: cycles of expansion, contraction, and consolidation

A line that stuck

“You don’t win by being another tool—you win by defining the category everyone else has to fit into.”

Call to Action

If you enjoyed this conversation, share it with a founder or tech leader who’s navigating the enterprise market. Make sure to follow the show for more unfiltered conversations with people shaping the future of software and AI.

Jonathan DiVincenzo, co-founder and CEO of Impart Security, joins the show to unpack one of the fastest growing risks in tech today: how AI is reshaping the attack surface. From prompt injections to invisible character exploits hidden inside emojis, JD explains why security leaders can’t afford to treat AI as “just another tool.” If you’re an engineering or security leader navigating AI adoption, this conversation breaks down what’s hype, what’s real, and where the biggest blind spots lie.

Key Takeaways

• Attackers are now using LLMs to outpace traditional defenses, turning old threats like SQL injection into live problems again

• The attack surface is “iterating,” with new vectors like emoji-based smuggling exposing unseen vulnerabilities

• Frameworks have not caught up. While OWASP has listed LLM threats, practical solutions are still undefined

• The biggest divide in AI coding is between senior engineers who can validate outputs and junior developers who may lack that context

• Security tools must evolve quickly, but rollout cannot create performance hits or damage business systems

Timestamped Highlights

01:44 Why runtime security has always mattered and why APIs were not enough

04:00 How attackers use LLMs to regenerate and adapt attacks in real time

06:59 Proof of concept vs. security and why both must be treated as first priorities

09:14 The rise of “emoji smuggling” and why hidden characters create a Trojan horse effect

13:24 Iterating attack surfaces and why patches are no longer enough in the AI era

20:29 Is AI really writing production code and what risks does that create

A thought worth holding onto

“AI is great, but the bad actors can use AI too, and they are.”

Call to Action

If this episode gave you new perspective on AI security, share it with a colleague who needs to hear it. Follow the show for more conversations with the leaders shaping the future of tech.

Daniel Saks, co-founder and CEO of Landbase, joins The Tech Trek to unpack the real meaning of democratizing technology. From agentic AI that works for you—not the other way around—to rethinking workflows and change management, Daniel shares why this shift is bigger than the move from on-prem to cloud. For tech leaders, founders, and operators, this episode reveals how to reclaim time, scale smarter, and prepare for the next wave of AI-native business.

Key Takeaways

• AI is moving beyond hype—it’s becoming the engine that executes real workflows and shifts power from systems to users

• Businesses that recapture saved time will unlock significant cost efficiency and growth potential

• The gap between idea and implementation is shrinking fast, but durable value will come from solving the hardest problems, not the easiest apps

• Change management is now about building AI-native workflows and cross-functional systems, not just adopting tools

• Sales and go-to-market leaders can gain an edge by mastering prompting and AI-driven enrichment today

Timestamped Highlights

00:56 — Why Landbase built GTM-1 Omni to reimagine go-to-market execution

01:40 — From on-prem to cloud to AI-native: the next major leap in democratizing technology

04:34 — Why fears about AI replacing jobs miss the bigger story of new roles and industries emerging

08:42 — How the pace of product cycles is collapsing and what that means for value creation

13:25 — Inside Landbase’s “AI Factory” model for automating workflows across functions

16:39 — What people actually do with the time they reclaim through AI-driven automation

19:23 — How AI is reshaping the role of the salesperson and why adoption speed matters

A line that stood out

“You don’t have to work for your software anymore—your software works for you.”

Call to Action

If this conversation gave you fresh ideas about how AI is reshaping business, share it with your team and subscribe to The Tech Trek on Apple Podcasts or Spotify. For more insights, follow along on LinkedIn.

Matt McLarty, CTO at Boomi, joins the show to break down what enterprise AI adoption really looks like in 2025. From navigating the hype cycle to identifying practical first steps, Matt shares what separates companies that are seeing value from those stuck in endless pilots. If you’re a tech leader wondering how to move beyond experimentation and into measurable outcomes, this episode is your playbook.

Key Takeaways

• AI adoption is not binary—it’s a spectrum, and success depends on linking it to business value, not just “using AI.”

• Orientation matters: every company needs an honest assessment of where they are on their digital maturity curve before jumping in.

• Small, low-risk bets build the organizational muscle memory required for bigger wins.

• The fastest wins often come from augmenting existing automation rather than chasing moonshots.

• Companies that succeed treat AI as a tool to solve business problems, not as an end goal.

Timestamped Highlights

01:38 – Why AI’s hype cycle feels like “Mount Everest” compared to cloud and mobile

04:50 – Why AI adoption can’t be compared to past waves like blockchain or cloud

07:36 – The hidden foundation: digital transformation work still matters

11:11 – The inversion that changes everything: AI isn’t the goal, business outcomes are

16:26 – Defining “adoption” as a multi-dimensional spectrum, not a checkbox

19:50 – How to recover if your first AI projects fall short

28:04 – Building adaptability as a core enterprise competency

31:25 – The common traits of companies succeeding with AI right now

A standout moment

“AI isn’t the end goal—it’s just another tool. The real question is, what business problems can we finally solve with it?” – Matt McLarty

Call to action

If this episode gave you a clearer path toward enterprise AI adoption, share it with a colleague and follow the show so you never miss a conversation on where tech leadership is heading.