Engineering Choices You Have to Defend

About Engineering Choices You Have to Defend

Business leaders are being told they need AI. They are rarely told what happens when it goes wrong. Each episode is a conversation with an owner or executive about a real decision they made about AI at their organization: what they adopted, what they held off on, what it cost them, and what they wish someone had told them first. These are leaders who own the AI decision but not the engineering, and most of them are working it out without a technical team behind them. No jargon and no vendor pitches. Just the decisions and what came of them. Hosted by Nicola Onassis of Code Particle.

Publisher

Nicola Onassis

Category

education · news · technology

Language

en

Explicit

No

First episode

3 Mar 2026

Latest episode

9 Oct 2026

Latest episodes

23 episodes in the feed.

1. 

   9 Oct 2026

   How Anya Gerasimchuk Uses AI to Blur the Line Between UX Design and Engineering

   Episode Summary: In this episode of Engineering Choices You Have to Defend, hosted by Nicole Onassis, we sit down with Anya Gerasimchuk, a product and UX leader who has built digital products and led global teams. Anya shares how AI is changing the way design teams build software, develop design systems, and collaborate with engineering. The conversation explores how a design system originally created for a patient-facing oncology portal struggled to support the growing needs of electronic health records and life sciences products. These applications required more data-rich interfaces for doctors, nurses, and other healthcare professionals. With limited engineering resources, Anya’s team began using AI to generate components and turn requirements into working code, reducing a process that once took six months to just a few months. Anya explains how AI helps designers gather requirements, build components, and automate updates to design foundations more efficiently. She also emphasizes that AI-generated code still requires careful testing, accessibility considerations, and collaboration with engineering, quality assurance, and product teams to ensure components work reliably in production. The discussion also explores how AI is changing traditional boundaries between design and engineering. While designers can now take greater ownership of implementation, engineering remains essential for infrastructure, data management, performance, and backend decisions. Anya highlights why user research and understanding customer needs remain critical, even as AI makes software development faster. Key Takeaways: AI allows UX teams to build design system components faster and take greater ownership of implementation Clear requirements and strong technical foundations help AI generate more reliable code Testing, accessibility, maintainability, and collaboration remain essential for production-ready components AI is changing the boundaries between design and engineering, but it does not eliminate the need for engineering expertise Faster development creates more opportunities for user research and a deeper understanding of customer needs Connect with Anya Gerasimchuk: LinkedIn: (https://www.linkedin.com/in/avgera/)https://www.linkedin.com/in/avgera/ (https://www.linkedin.com/in/avgera/) Website: (https://anyagerasimchuk.com/)https://anyagerasimchuk.com/ (https://anyagerasimchuk.com/) Listen Now & Subscribe: Apple Podcasts, Spotify, Amazon Music, or wherever you get your podcasts. "Engineering Choices You Have to Defend explores the real technical decisions behind regulated software, engineering transformation, AI integration, and the systems leaders build to make complex technology work in the real world.

2. 

   29 Sept 2026

   How John Woodyard Proves Developer Productivity Can Outperform a $20 Million Optimization

   Episode Summary: In this episode of Engineering Choices You Have to Defend**, hosted by Nicola Onassis sits down with John Woodyard, an experienced software engineering and DevOps leader with more than 25 years of experience. John shares a lesson from his time working on service frameworks at Amazon, where the team initially identified an opportunity to save roughly $20 million through infrastructure performance improvements. The challenge was that not all services benefited equally from the same optimization. With tens of thousands of microservices serving different workloads, performance improvements that made a major difference for large services like DynamoDB did not necessarily translate across the broader ecosystem. John explains how the team realized that developer productivity could represent a much larger opportunity by reducing the repetitive work required to build and launch services. The conversation explores how John and his team measured developer productivity by identifying common tasks engineers performed across most services, calculating the time spent on that work, and translating those savings into business value. One example involved authentication, authorization, and auditing, where a process that previously took roughly two weeks was reduced to essentially uncommenting a line of code. John explains how that change alone contributed to more than $2 million in savings in the first three months. John also discusses how AI is changing the definition of developer productivity. As AI makes it possible to generate code faster, the bottleneck can move upstream toward deciding what should be built and downstream toward security, testing, scanning, and software delivery. For engineering leaders, John emphasizes that the right optimization depends on the business constraint: sometimes performance matters most, while in other situations removing repetitive engineering work creates greater impact. Key Takeaways: A large, measurable optimization can still be the wrong engineering priority Developer productivity can be measured through time savings, engineering costs, and service creation data Removing repetitive work allows engineers to focus more on innovation and business logic AI may shift software development bottlenecks toward requirements, security, testing, and delivery Engineering investments should be tied to the business constraint they are designed to solve Connect with John Woodyard: LinkedIn: (https://www.linkedin.com/in/john-woodyard/?utm_source=chatgpt.com)https://www.linkedin.com/in/john-woodyard/ (https://www.linkedin.com/in/john-woodyard/?utm_source=chatgpt.com) Listen Now & Subscribe: Apple Podcasts, Spotify, Amazon Music, or wherever you get your podcasts. "Engineering Choices You Have to Defend explores the real technical decisions behind regulated software, engineering transformation, AI integration, and the systems leaders build to make complex technology work in the real world."

3. 

   23 Sept 2026

   How Manju Abraham Turns Engineering Transformation Into a Systematic Path to Production

   Episode Summary: In this episode of Engineering Choices You Have to Defend, hosted by Nicola Onassis sits down with Manju Abraham, a senior engineering executive and former VP of Engineering. Manju shares how she has led large global engineering organizations through complex transformations where reliability, quality, execution, and customer needs all have to stay aligned. The conversation explores what happens when engineering organizations inherit products through multiple acquisitions, leaving teams with different code bases, platforms, release processes, tools, and definitions of quality. Manju explains why transformation starts with understanding the root cause, using data and metrics to identify systemic problems such as unstable infrastructure, unreliable testing, and broken CI/CD processes. Manju also discusses the challenge of balancing legacy products with new platforms. While engineering teams need to build for the future, existing customers still depend on products that generate revenue today. She explains how a clear North Star, strong collaboration between engineering and product teams, and disciplined sequencing can help organizations make progress without sacrificing stability. The conversation also explores AI transformation and the difficult transition from proof of concept to production. Manju explains why successful AI systems require more than a working model. They need clear ownership, budgets, evaluation frameworks, release processes, on-call responsibilities, and accountability for how AI agents behave. She also emphasizes the importance of communication, training, and trust as organizations redesign roles and workflows around AI. For engineering leaders, Manju argues that lasting transformation is less about adopting the latest technology and more about strengthening the people, processes, and foundations that allow technology to scale. Her goal is to be a “fire preventer,” building systems that address recurring problems at the root rather than repeatedly reacting to the same failures. Key Takeaways: Root-cause analysis helps engineering teams solve systemic problems instead of recurring symptoms Data and clear metrics create a shared understanding of engineering challenges Legacy products require stability and customer support while new platforms are being built AI proof of concepts need ownership, evaluation, accountability, and production processes to scale Successful transformation depends on people, processes, communication, and trust as much as technology Connect with Manju Abraham: LinkedIn: (https://www.linkedin.com/in/manju-abraham)https://www.linkedin.com/in/manju-abraham (https://www.linkedin.com/in/manju-abraham) Email: manju.abraham@gmail.com (mailto:manju.abraham@gmail.com) Listen Now & Subscribe: Apple Podcasts, Spotify, Amazon Music, or wherever you get your podcasts. "Engineering Choices You Have to Defend explores the real technical decisions behind regulated software, engineering transformation, AI integration, and the systems leaders build to make complex technology work in the real world."

Reach and audience

Public platform figures. Ratings count people who left a rating, not total listeners.

Apple Podcasts (US)

5.0 / 5

1 ratings

Contact Engineering Choices You Have to Defend

Guest appearances

Books guests

Based on episode analysis; this does not confirm that the show is currently accepting guests.