Frontend CI/CD in the age of AI - Part 2: Deployments
Agents write code that passes every test and still breaks in production. Here is how to ship it to a slice of your users first, and how to take it back in seconds when it goes wrong.
Neciu Dan is a seasoned software engineer with over 14 years of expertise in the tech industry, host of the Señors @ Scale podcast, and creator of the sold-out Lizard to Wizard engineering workshop. He has a proven track record, having served as the technical co-founder and tech lead of CareerOS, a Senior Product Engineer at the New York e-commerce brand AdoreMe, and a Senior Software Engineer at the food delivery company Glovo.
Beyond his technical contributions, Dan is an esteemed international speaker and mentor. He passionately encourages best development practices, with a keen focus on testing, accessibility, and security. His recent speaking engagements include esteemed conferences like React Alicante, React Summit, React Rally, Utah JS, Techorama, and Frontmania.
Discover actionable insights from conversations with senior engineers about scaling applications, teams, and careers.
Señors @ Scale host Neciu Dan sits down with Craig Abbott, Principal Accessibility Specialist at TetraLogical and the former Head of Accessibility at the UK's Department for Work and Pensions, one of the largest government departments in the country, where he built a dedicated accessibility practice from nothing and open sourced the DWP Accessibility Manual. Craig has over 15 years in user centred design and has led accessibility work across the public sector and at Elastic. From what sustainable accessibility actually means and why third party audits alone don't get you there, to the three C's of compliance, culture and capability, to running screen readers in VMs without expensive licences, to accessibility acceptance tests in CI with Playwright, Cucumber and Guidepup, to why compliance does not mean usable, this is the accessibility conversation for teams who want it to survive the person who cares about it.
Señors @ Scale host Neciu Dan sits down with Carmen Huidobro, CTO at Incredible Bee in Vienna, where she builds products and helps teams figure out what should be automated and what should stay in the hands of users. Carmen has spent 17 years in tech, almost all of it freelancing, working across Objective-C, Ruby on Rails, the web, mobile, hardware, and even ABAP inside an SAP consultancy, plus five years in developer relations and developer education. Her argument is that the generalist versus specialist debate misses the point: the durable skill is being an expert at adapting. From adding a local Mistral model to a twenty-year-old macOS app without betraying the people who use it, to the refugee hackathon project the City of Vienna still runs a decade later, to why she won't take money from junior developers, this is a conversation about the skills that survive the shift.
Señors @ Scale host Neciu Dan sits down with Marko Gaćeša, Head of Product at Semaphore, the agent-native CI/CD platform, and the first product guest on the show. Marko is a serial entrepreneur whose career spans developer tools, IoT, industrial automation and enterprise SaaS, including founding Dry Tools and serving as Chief Product Officer at Alchemy Cloud, with earlier work in domains like medical equipment where quality is non-negotiable. From why testing rather than coding is now the bottleneck, to what agent-native CI/CD actually means when developers live inside their coding agent, to how SemAI attacks flaky tests and automates migration off GitHub Actions, to pricing a platform where a minute of CI is not the same minute everywhere, this is the product side of developer tooling from someone shipping it.
Señors @ Scale host Neciu Dan sits down with Maxim Salnikov, AI Dev Tools Solution Engineer at Microsoft, where he leads AI native development enablement for over 100 enterprise customers of Microsoft and GitHub in Norway. Maxim has been building for the web since the late 90s and spends his days inside real enterprise dev teams across finance, energy, agriculture, and pure software companies, watching AI adoption succeed and fail. From why adoption is a change management problem rather than a technology one, to the anatomy of an AI harness and the external layer successful companies build on top of it, to the context engineer and agent ops roles now appearing in team topologies, to managing agent skills as versioned dependencies instead of letting them pollute the repo, this is the enterprise AI adoption conversation from someone who sees a hundred versions of it.
Señors @ Scale host Neciu Dan sits down with François Best, freelance web developer, founder of 47ng, and the creator of nuqs — the type-safe search param state manager for React with over 10,000 GitHub stars, used by Sentry, Supabase, Vercel and Clerk. François started nuqs during the pandemic to move a woodworking calculator's state from his laptop to his phone, left it sitting on GitHub for two years, and rewrote it into a framework-agnostic library after Next.js shipped the app router. From the async scheduling queue that works around browser History API rate limits, to render count budgets enforced in CI, to the poisoned GitHub Actions cache behind the TanStack compromise, to running open weight models locally for clients who can't send code across the Atlantic, this is the URL-as-state conversation from the person who had to make it actually work.
Señors @ Scale host Neciu Dan sits down with Susanna Wong, staff software engineer at Typeform and a Google Developer Expert in Web Technologies since 2020. Susanna's path is anything but typical: she built her own major at the University of Virginia, wrote software prototypes at SAP Labs in Palo Alto, then spent close to seven years as a building architect and computational designer, writing scripts that optimized facade panels at Arup. She came back to software in 2016, built Kedro and Kedro-Viz at QuantumBlack during the MLOps boom, and now works on AI features and a multi-agent system at Typeform. From why the LLM replaced the trained model as the baseline, to when a browser-hosted model beats an API call, to running offline and online evals against a non-deterministic feature, to how WebMCP stops agents from doing OCR on your login page, this is the frontend AI conversation grounded in someone shipping it.
Señors @ Scale host Neciu Dan sits down with Sébastien Lorber, creator of This Week in React and a maintainer of Docusaurus at Meta. Sébastien has worked with React since late 2013 and spent years as an independent consultant before a French-language side project aimed at landing audit work turned into the newsletter 45,000 developers open every week. From the exhaustive weekly workflow across 2,000 X profiles and 500-plus RSS feeds, to the Gmail size limit that quietly decides what makes each issue, to why French sponsors could never make it sustainable, to why he checks React's feature flags before telling anyone a feature has shipped, this is the unglamorous operational reality behind one of the biggest newsletters in the React ecosystem.
Señors @ Scale host Neciu Dan sits down with Santosh Yadav, principal developer advocate at CodeRabbit and one of only around 80 GitHub Stars in the world. Santosh started hating C in 2004, fell for C# by 2008, and turned a year of open source contributions to Angular and NgRx into a stack of community titles — Google Developer Expert, GitHub Star, Nx champion, and Microsoft MVP. As a staff engineer at Celonis he led the move of 20-plus apps to module federation and drove Nx adoption across 30-plus teams when the product grew from four apps to thirty. From the year-long incremental migration off a single deployable unit, to why polyrepos can't give AI tools the context they need, to how Nx's affected graph and build caching tame a 20-million-line monorepo, to running code review for free for open source at CodeRabbit, this is the monorepo conversation grounded in someone who actually shipped one at scale.
Agents write code that passes every test and still breaks in production. Here is how to ship it to a slice of your users first, and how to take it back in seconds when it goes wrong.
Code is written faster than ever with AI loving to write tests for everything it does. This makes our CI pipeline slower and slower. Here is how to build a fast CI/CD pipeline - Part 1 with focus on CI
Let's find out by exploring how the most popular libraries are doing it under the hood and then build our own.
The React Compiler memoizes your components at build time, and it does the job better than you can by hand. What it is, how the compiled output works, how to turn it on, and what to do with the useMemo calls already in your codebase.
Where should your auth token live so an XSS bug can't steal it? Here's how to build auth that survives the crazy non-secure world we live in.
Over the years, in our goal to achieve faster and faster web applications, we created different hydration and rendering strategies. Each with benefits and drawbacks that we explore in this article.
Check out my latest educational videos on React, JavaScript, and frontend development best practices.
View all videosLearn the keys to success in web development, from accessibility to performance, and how to craft a secure software