Advertise on podcast: The Tech Trek

The original idea may not be what makes a startup successful. As AI changes markets and creates new opportunities, the ability to learn and change direction becomes increasingly important. Tim Chen, Managing Partner at Essence Venture Capital, invests in highly technical founders building infrastructure and developer tools. A former engineer and founder himself, Tim started his fund to provide the kind of support he wished he had when raising capital. In this episode, Tim explains why his investment approach has evolved as AI accelerates change. He shares how he evaluates technical teams, what makes a pivot worth pursuing, and why he believes specialization helps investors stand out. The conversation also explores his decision to operate as a largely solo investor, the importance of founder autonomy, and what it means to build a venture fund around a specific area of expertise. Key Takeaways Why a team's ability to learn matters beyond its initial product idea. How AI is changing investment decisions in infrastructure and developer tools. Why specialist investors can offer founders more relevant connections and guidance. How Tim approaches pivots without making decisions for founders. Episode Highlights 00:30 Why Tim started Essence Venture Capital. 03:32 The advantages of making independent investment decisions. 12:08 How specialist investors demonstrate value before investing. 17:03 What the Tabular investment revealed about technical opportunities. 18:04 Why some of Tim's strongest investments have involved major pivots. 22:33 How Tim evaluates the reasoning behind a founder's decision to pivot. A key quote: "Don't be too religious about the approach you take because you most likely are wrong." Follow The Tech Trek for more conversations about engineering, AI, startups, and technical leadership.

An AI agent can be fast, available, and still get the answer wrong. That changes what engineering teams need to measure. Jeetendra Soneja, VP of Engineering, AI Platform at iHerb, joins The Tech Trek to discuss why AI observability goes beyond traditional software monitoring. Teams need to understand not just latency and errors, but also response quality, groundedness, cost, and whether an agent completes its intended task. We also explore how to measure AI ROI without relying on misleading productivity metrics. Jeetendra shares his perspective on risk based guardrails, the skills engineers need, and why observability should begin during development rather than waiting until production. Key Takeaways Monitor agent quality and task completion alongside latency, availability, and cost. Build evaluations around the specific business problem the AI system is solving. Connect AI spending to business outcomes, not just token usage or code output. Start with basic observability, then expand controls as AI adoption grows. Episode Highlights 01:00 Why AI observability requires more than traditional monitoring. 06:00 The metrics that matter for cost, quality, and reliability. 08:00 How to measure AI ROI beyond productivity claims. 11:00 Setting risk tolerances and choosing appropriate guardrails. 16:00 The engineering skills becoming more valuable with AI. 22:00 Proving value before scaling AI systems. One Line That Stuck "It's one thing to get it done, it's another thing to get it done economically." Follow The Tech Trek for more conversations about building and operating modern technology teams.

What happens to engineering teams when everyone can code? Bali Raghavan, cofounder and CTO at Hero Safety, joins Amir Bormand to discuss how AI is changing team structures, engineering skills, and hiring. At Hero Safety, designers, product managers, and operations teams are using AI to build and solve problems that once required engineering support. That shift changes what companies need from engineers. Bali explains why systems thinking, architecture, judgment, and problem solving are becoming more valuable. He also shares why he now expects engineering candidates to use AI during interviews. The conversation explores where specialized expertise remains essential, including computer vision and edge AI, and how public safety organizations are approaching AI adoption with both interest and skepticism. Key Takeaways AI is expanding who can build software across a company. Engineering judgment and architecture matter more as code generation becomes easier. Coding interviews should reflect how engineers actually use AI at work. Specialized skills remain essential for complex AI systems and edge computing. Episode Highlights 02:15 | Why Bali considers his entire company an engineering team 04:00 | How AI changes the value of programming skills 08:30 | Judgment, first principles, and evaluating AI output 11:45 | Why using AI is now expected in engineering interviews 14:25 | Generalists, specialists, and the need for edge AI expertise 16:15 | AI adoption, privacy, and skepticism in public safety One Line That Stuck "The bug has become a feature." Bali explains how AI use went from something employers tried to prevent in interviews to something they now want to evaluate. Follow The Tech Trek for more conversations about engineering leadership, AI, and building technology companies.

Getting a voice agent to talk is one challenge. Knowing whether it can handle real conversations is another. Brooke Hopkins, founder of Coval, joins The Tech Trek to discuss how teams test, evaluate, and scale voice AI agents. Drawing on her experience leading evaluation infrastructure at Waymo, Brooke explains why testing voice agents shares surprising similarities with testing autonomous vehicles. The conversation explores why older voice assistants struggled, how better reasoning models changed what's possible, and what it takes to earn users' trust. Brooke also shares how her engineering team works with multiple AI agents, why technical interviews need to assess AI skills, and how she balances automation with human judgment as a founder. Key Takeaways - Voice AI evaluations must account for interruptions, background noise, and complex conversations. - Simulations identify failures before deployment, while production QA reveals what testing missed. - Engineers need new ways to divide work when running several AI agents in parallel. - Hiring should assess problem solving, system design, and how candidates use AI. Episode Highlights 00:32 How Coval tests voice agents before and after deployment 01:26 What autonomous vehicle simulation teaches us about voice AI 03:24 Why better speech recognition wasn't enough 08:00 Could voice AI change how we use screens? 14:43 How AI is changing engineering interviews 16:43 Where a founder uses AI and where human judgment matters A Moment Worth Pulling Out "We had the eyes and the mouth, but we didn't have the brain." Follow The Tech Trek for more conversations about engineering, AI, and building technology companies.

What happens when AI changes more than coding speed? Chris Merrick, Co Founder and CTO at Omni, explains how his engineering team is adapting around AI, autonomy, product ownership, and a lighter operating model. Omni has about 25 engineers working in groups as small as one to three people. Engineers are expected to make product decisions, move without waiting for detailed specifications, and show unfinished work during a weekly company demo day. Those demos are also posted publicly. Customers watch them, join betas, and see features often move into production soon after. Chris also shares how Claude Code changed development across the team. Commits to main increased roughly two to three times without comparable team growth, but the added output created another constraint: code review. Takeaways • Small teams can operate with less process when engineers have real ownership. • Public demos create feedback loops between engineers, customers, and the rest of the company. • AI can increase engineering output while moving the bottleneck somewhere else. • Token costs still matter when AI usage scales across products and engineering teams. Key Moments 01:15 Building AI into the engineering workflow 03:38 Why Omni works in teams of one to three engineers 06:48 The weekly demo day that replaces heavier process 10:29 Why Omni publishes unfinished product demos publicly 17:14 Measuring AI engineering productivity through commits to main 19:52 When token economics start affecting engineering decisions One Line That Stuck “I’m not married to the process. I’m married to the outcomes.” Follow The Tech Trek for more conversations about how modern engineering and product teams are changing.

AI coding tools are helping engineers write code faster. But faster code does not automatically mean better business outcomes. Vitaly Gordon, cofounder and CEO of Faros, joins The Tech Trek to talk about the gap between AI adoption and measurable results. The conversation looks at why engineering teams are spending more on AI, how enterprises are thinking about ROI, and why existing processes can become the bottleneck even when code generation speeds up. Vitaly also explains why AI adoption inside large organizations looks different from simply giving engineers access to new tools. Teams still have to deal with security, legacy processes, risk, integration, and organizational change. Key takeaways • More AI usage does not automatically create more customer value. • Engineering leaders need to connect AI spend to actual outcomes. • Enterprise adoption requires process changes, not just new tools. • AI ready engineers increasingly need both domain expertise and AI skills. Key moments 00:36 Stop token maxing and start outcome maxing 02:51 Why FOMO accelerated AI adoption 06:32 Measuring ROI on engineering AI spend 09:19 Change management inside large engineering organizations 14:23 Why AI ready engineers are harder to hire 17:13 Will AI reduce engineering jobs? Best Line “Stop token maxing and start outcome maxing.” Follow The Tech Trek for more conversations with technical leaders building and operating modern teams.

Technology teams are often asked to solve something before everyone agrees on what the actual problem is. Chad Carrington, CIO at Golden1 Credit Union, joins The Tech Trek to talk about listening to business partners, defining the outcome first, and resisting the instinct to jump immediately to a technology solution. The conversation also looks at how AI is changing this skill. As tools make building easier and faster, people still need to articulate what they actually want, ask better questions, and understand what success looks like. What You Will Hear • Why listening and hearing are not the same thing • When solving 85 percent of a problem may be enough • Why technology teams need more time with business users • How AI is increasing the value of asking better questions Key Moments 01:20 Listening versus actually hearing the business problem 03:47 When immediate customer impact matters more than technical debt 06:31 Bringing the business into decisions about speed and tradeoffs 09:47 Why technologists need to spend more time with business teams 13:20 How easier tools change the importance of defining the problem 18:33 Why AI is still a tool, not a substitute for expertise One Line That Stuck “There’s a difference between listening and hearing.” Follow The Tech Trek for more conversations with the people building and leading modern technology teams.

AI has learned from the digital world. Physical AI brings real world data into the picture. Doron Hazan, Director of Products and AI at Wiliot, joins The Tech Trek to explain how physical AI connects AI systems with objects, environments, and supply chains. The challenge is not simply processing data. It is collecting accurate, current information from the physical world. Doron explains how ambient IoT, sensors, statistical inference, and cloud systems can help companies understand where assets are, what condition they are in, and what may happen next. The conversation also covers the role of human judgment. Supply chains require many decisions, often with consequences that spread across the system. That makes guardrails and human involvement especially important. Key Takeaways • Physical AI connects AI systems with data from the real world. • Better supply chain visibility starts with accurate, current physical data. • Real time decisions matter, but decision accuracy matters more. • Human judgment remains important when AI affects physical operations. Highlights 01:53 What separates physical AI from traditional AI 03:18 Why real world data collection changes the problem 07:19 Supply chain visibility and practical use cases 11:57 How quickly physical AI systems can make decisions 12:48 Why guardrails matter in supply chain automation 15:28 Robotics, distributed physical AI, and connected systems Follow The Tech Trek for more conversations about AI, engineering, product, and technical leadership.

Scott Hanson went straight from a PhD program at the University of Michigan to building a semiconductor company. Today, as Founder and CTO of Ambiq, he is working on the same core idea that inspired the company years ago: putting intelligence into the devices around us. What changed is what those devices can now do. Scott shares what it was like becoming CEO without prior industry experience, why moving into the CTO role was harder than expected, and how the rise of AI accelerated Ambiq’s original vision. The conversation also looks at what happens as more AI processing moves closer to the device, from wearables and smart homes to factories, medical devices, infrastructure, and smart glasses. Key Takeaways • Founder roles may need to change as the company grows. • Edge AI can reduce how much personal data needs to leave a device • Low power computing expands where AI can operate. • AI tools are changing engineering work from coding toward architecture and design. Key Moments 01:57 Going directly from a PhD program into a startup 06:42 Why Scott moved from CEO to CTO 10:40 Smart dust and Ambiq’s original vision 13:43 Why AI is moving beyond the cloud 17:52 Privacy, security, and processing data locally 20:04 Industrial, medical, and smart glasses use cases A Moment Worth Pulling Out “Be present where your feet are.” Follow The Tech Trek for more conversations with the people building and operating modern technology companies.

Building an AI product is getting easier. Building an AI company that lasts is not. Itamar Novick, Founder and General Partner at Recursive Ventures, joins The Tech Trek to explain what he looks for when investing at the earliest stages of AI companies. The conversation covers how lower development costs could change venture funding, why subject matter expertise matters more as software becomes easier to build, and what actually creates defensibility when competitors can move quickly. Itamar also shares how Recursive Ventures thinks about founder anti patterns. Rather than trying to copy the paths of successful startups, he argues that founders can improve their odds by recognizing common mistakes that repeatedly create unnecessary risk. Key Takeaways • AI may let companies reach scale with much less outside capital. • Subject matter expertise matters more when building software is no longer the main barrier. • Proprietary data, feedback loops, hardware, and exclusive access can create stronger moats. • Founders can reduce risk by learning to recognize repeatable startup mistakes. Episode Highlights 00:38 What Recursive Ventures looks for in early AI companies 05:42 How AI could change the amount of capital startups need 10:04 Why subject matter expertise is becoming more valuable 12:02 What creates an AI moat when software is easy to copy 17:47 Why studying failure can be more useful than copying success 23:13 How AI could reshape venture investing itself Follow The Tech Trek for more conversations with founders, investors, and technology leaders building what comes next.

AI coding agents can produce software faster, but they do not replace the judgment needed to understand the system. Shaun Patterson, CTO at Titan, joins The Tech Trek to discuss how agentic coding is changing problem solving, development workflows, project management, and technical hiring. Shaun explains why engineers still need a strong mental model of the systems they are building. AI can generate code, reproduce bugs, research implementation options, and automate repeated debugging work. But it can also keep working on the wrong problem long after a human debugger would have found the answer. The conversation also gets into a bigger shift in software delivery. If agents can work across much larger pieces of a project, engineering teams may move from managing work at the story level to working at the epic level. Key Takeaways • AI speeds up implementation, but engineering judgment still matters. • Repeated debugging work can become reusable agent skills. • Faster implementation lowers the cost of testing different technical approaches. • Hiring increasingly needs to measure how engineers work with AI. Highlights 02:08 Why AI can abstract work, but not engineering wisdom 06:04 Turning repeated debugging sessions into reusable agent skills 09:47 Why faster development may change traditional project management 12:42 Moving engineering work from stories to epics 16:19 Where agentic coding still creates problems 19:29 How Titan evaluates engineers who use AI One Line That Stuck “It abstracts your thinking, but it doesn’t abstract your wisdom.” Follow The Tech Trek for more conversations with the people building and leading technology companies.

AI agents create a different security problem from traditional software. They can operate at software speed and scale while behaving in ways that are much less predictable. Ev Kontsevoy, CEO and cofounder of Teleport, joins The Tech Trek to discuss what happens when companies deploy agents into security systems designed around humans, applications, and relatively static organizational structures. The conversation gets into authentication, impersonation, infrastructure identity, access control, and a harder question: what actually defines the identity of an AI agent when its model, memory, skills, and capabilities can change? Ev also explains why the combination of speed, scale, and unpredictable behavior changes the risk of mistakes. Later, he explores the tension between agents being useful because they can do new things and security systems that often depend on predictable behavior. Key takeaways • Agent identity gets harder when memory, models, and capabilities can change. • Traditional access controls often reflect static organizational structures. • Agents combine software speed with behavior that can be difficult to predict. • Useful agent behavior can conflict with security systems built around anomaly detection. Highlights 00:41 What Teleport does and why infrastructure identity matters 08:37 Why companies may already be behind on agent security 13:47 Why an electronic account is not the same as identity 15:11 What actually defines the identity of an AI agent? 22:49 Why agent speed and unpredictability change the risk equation 29:04 The conflict between useful agent behavior and anomaly detection One Line That Stuck “Agents are just as unpredictable as humans, but they are way, way, way faster.” Follow The Tech Trek for more conversations with the people building and leading technology companies.

AI agents are moving beyond helping with individual tasks. The bigger question is how much of a company they can actually run. Ben Cera, founder of Polsia, joins The Tech Trek to discuss what happens when AI handles engineering, support, marketing, research, and other parts of company execution. Ben explains how Polsia uses specialized agents that can take direction from a founder or decide what to work on autonomously. He also shares how he uses similar systems inside his own company, which he says has more than 10,000 paying customers and is approaching a $10 million run rate without a traditional full time team. The conversation gets into where humans still matter, why AI mistakes may be acceptable, and how faster execution changes the way founders test ideas. Key Takeaways • AI agents can move from completing tasks to coordinating entire business functions. • Faster execution gives founders quicker feedback on what works and what does not. • Humans still matter most for judgment, direction, and authentic storytelling. • Autonomy requires accepting some mistakes instead of demanding perfect AI output. Highlights 02:43 What changes when AI becomes part of how a founder operates 04:03 Turning customer support into a system that can also fix problems 06:19 Running a company without a traditional full time team 13:47 Why founder judgment still matters when AI gives the options 18:49 How specialized agents coordinate engineering, marketing, and outreach 22:10 What happens when autonomous AI makes the wrong decision One Line That Stuck “You have to trust your gut and you have to be willing to make mistakes.” Follow The Tech Trek for more conversations with the people building and leading technology companies.

AI coding agents can help engineering teams ship more code. But the bigger change may be what engineers spend their time doing. Viren Baraiya, Co-Founder and CTO of Orkes, joins The Tech Trek to discuss how AI is changing workflow orchestration, engineering productivity, project delivery, and hiring. As agents take on more implementation work, engineers are spending more time on design, architecture, review, and verification. Viren shares how his team measures the return on AI through product velocity, stability, and the ability to build things that previously required more time or outside resources. He also explains how Orkes manages model costs by using stronger models for difficult reasoning and smaller models for implementation. Key Takeaways • Coding agents increase output, but they also increase the need for verification. • Engineers are shifting from pure implementation toward design, review, and orchestration. • Repeated AI tasks can become reusable workflows that reduce ongoing token usage. • Hiring should test how engineers actually work with agents, not just manual coding. Highlights 01:21 Why agents are workflows and where orchestration fits into AI systems 05:19 How AI changed feature velocity, testing, and customer engineering at Orkes 07:44 Measuring AI ROI through velocity, stability, and new product capabilities 09:19 Why engineers increasingly look more like tech leads 12:56 Turning repeated AI requests into reusable workflows to reduce token usage 19:34 Why Orkes changed engineering interviews to include agentic coding One Line That Stuck “That has become a more important skill than actually writing the code now.” Follow The Tech Trek for more conversations with the people building and leading technology companies.

AI is doing more than helping engineers write code faster. It is starting to change who can participate in software development, how teams divide work, and where technical talent creates the most value. Shaosu Liu, Co Founder and CTO at Loop, explains how his team is using AI agents across the software development lifecycle while also enabling highly technical people outside traditional software engineering roles to build customer specific workflows. The result is a different model for scaling technical work, one that matters for founders and technical leaders thinking about team design in the age of AI. Loop is also using forward deployed engineers as core product engineers who can work directly with customers, understand difficult edge cases, and turn those requirements into product improvements. Takeaways • AI agents can now support much more than coding, including specifications, testing, deployment, validation, rollout, and production management. • Giving technical people better AI tools can expand who is capable of contributing to software development. • The final few percent of a customer workflow may consume most of the manual effort. AI makes that customization more practical. • Forward deployed engineers need both technical ability and the judgment, communication skills, and confidence to work directly with customers. Key Moments 04:37 How Loop uses AI across the software development lifecycle 07:05 Why people outside traditional engineering roles are writing significant amounts of code 09:03 Why automating the final few percent of a workflow can remove most of the remaining manual work 11:01 Why forward deployed engineers matter when customer requirements get complicated 16:37 Why forward deployed engineering is often a path to another role rather than a long term career 19:50 How Loop evaluates technical ability and customer facing skills when hiring One Line That Stuck “That last 5% automation ends up saving 100% of time.” Follow The Tech Trek for more conversations on AI, engineering, product, data, hiring, and technical leadership.