
Sign up to save your podcasts
Or


Alison Taffel Rabinowitz spent the first decade of her career in magazines, and the last twelve years teaching women how to negotiate via her company, The Finishing School. She'd never even thought to build software—until this year when she pulled on a hoodie and signed up for a class taught by comedian Alex Dobrenko. Her first project wasn't a productivity hack, but a dashboard to help a family member care for their partner with dementia. Since then she's built a horoscope site that pairs your daily reading with a mood board of maximalist outfits ("dopamine dressing"), a "headquarters" that rallied a team behind a nonprofit grant application, and a tour of her favorite New York boutiques guided by her dog, Typo. We get into why taste and wisdom are your late-career superpower, and why she tells her students to "make it till you make it." If you're skeptical about AI but curious anyway, Alison's joyful approach—and her upcoming AI Joy Jam for women—might be the invitation you need.
Links
Andy Balinsky spent his career in computer security—tracking intruders through military networks, back when that meant sleeping under a desk waiting for the alarm. Retired now, he's pointed that same attention at the things he loves: a podcast player that finally lets music keep playing underneath, a crib sheet for the hundreds of traditional Irish session tunes he's learned, and "Purple Skies", a field app for the Purple Martin colonies he and his wife have tended for 25 years. We get into why the challenging part of building software now isn't the code, but rather all the decisions that make it usable in the nebulous human realm. It's a whole new world when anyone can make helpful software like this—apps so specific, they certainly would never have been built and sold as commercial software. And yet, it means a lot to the people (and birds!) who benefit from it. It's an inspiring story, and a guide for anyone who wants to step up for their own interests and communities.
Links:
In today's episode, we speak with Michael Porcelli, founder of Metarelating, about the software he's been building for himself—not for other people, just for his own life. A former techie who left the field two decades ago, Michael walks us through three projects born out of ordinary domestic problems: a receipt splitter that divides a mile-long Costco bill item by item, a ledger that tracks his siblings' equity in an inherited house, and a database of his late father's 75,000-baseball card collection. Along the way we get into why he emphatically does not want to turn any of this into a product, the "four P's" framework for software (perishable, personal, public, and professional), and why having an AI agent is like having a personal handyman with a truck full of tools living in your back yard (and why sometimes, you still need to call a real structural engineer).In today's episode, we speak with Michael Porcelli, founder of Metarelating, about the software he's been building for himself—not for other people, just for his own life. A former techie who left the field two decades ago, Michael walks us through three projects born out of ordinary domestic problems: a receipt splitter that divides a mile-long Costco bill item by item, a ledger that tracks his siblings' equity in an inherited house, and a database of his late father's 75,000-baseball card collection. Along the way we get into why he emphatically does not want to turn any of this into a product, the "four P's" framework for software (perishable, personal, public, and professional), and why having an AI agent is like having a personal handyman with a truck full of tools living in your back yard (and why sometimes, you still need to call a real structural engineer).
Jonathan Milton has worked in healthcare operations for most of his career, wrestling with gnarly problems, like how to fill nursing shortages without relying too much on temp agencies. In this episode, we talk about how he turned his hard-won expertise into useful tools—like an online calculator that exposes the fully loaded cost of staffing, and a game to coach future health care execs on how to solve the human problems well. We dig into how removing the technical friction of building software lets someone like Jon—a domain expert, not a programmer—go from idea to working tool in an evening.
Jim Hanas spent his career in publishing—most recently as VP of Marketing at HarperCollins—but he hadn't been "under the hood" of technology in twenty years. Then a couple of friends got him "Claude-pilled", and within weeks he'd launched smallpressinsights.com, a dashboard that tracks the performance of small press books. In this conversation we get into his three principles for building (slow, thoughtful, clear), his "ocean and lenses" metaphor for handling data, and how he runs Claude Cowork as a PM that writes tickets for Claude Code to implement.
Links:
Today's guest is Mark Wochner: acoustic engineer, DIY aficionado, and maker of great cocktails. He got into AI app-building to revive a project he'd dreamed up called Historicast (look up weather in the past!) and got hooked enough to make himself a whole slew of other fun apps. In this episode, we talk about building software in hotel rooms across three continents, why he wants four weather forecasts instead of one, and why his one-person company finally got the part-time hire it needed (hint: it's Claude). Mark has used AI coding agents to make barcode apps and pantry databases, and in this conversation, we talk about why building software for an audience of one might be the most freeing thing you can do.
Links:
In today's episode, I sit down with Sharon Klardie—a longtime product leader who's spent her career around technology without ever being the one writing the code, and is now finally in the building seat thanks to what she calls her "AI minions." We talk about Board Forge, the app she built to make life easier for volunteer nonprofit boards (complete with an AI chief of staff named Jordan who tells you the one thing you actually need to do today), which just won "most improved" in week three of Replit's Buildathon. But it's not just one app—Sharon's also running an LLC that publishes puzzle books, building a math game to drill her 10-year-old's times tables, and inventing brand-new puzzle types on the side. Along the way she shares the practical stuff that's making it all work: treating her AIs like a team with a CEO and a QA engineer, having one model write prompts for another, telling them to stop brown-nosing and actually argue with her, and why the real currency she pays attention to is impact, not revenue. If you've ever had ideas for software that could help other people, but figured the tech part wasn't for you, this one will light a fire under you.
Links:
In this episode, I speak with the amazing Jacquelyn Halpern on how she got started with AI-assisted coding, and how she created a cooking app for herself and her family called Lil' Chef.
In this short introductory episode, I introduce what the Conjuring Code podcast is all about: helping regular people like you build software, using AI to help write the code. The pitch is simple: AI now handles all the tedious, mysterious "bad hard" parts of programming, leaving you free to focus on the "good hard" part of figuring out what's actually worth making.
From the publisher's feed
Conjuring Code is a podcast about building software, for non-software engineers.