Not long ago, building a functional app or tool required either a computer science degree or years of self-taught programming experience. Now? Coding know-how is no longer the barrier it once was.

With an AI assistant and the right prompt, anyone looking to build a simple app can probably get it done without writing a single line of code. This new kind of development is so prevalent that there’s a term for it: vibe-coding.

It works like this: In plain language, ask an AI assistant to build a tool, describing how you’d like that tool to work. For example, you could ask it to build you an app that sends you a daily verse from Tanakh, or a tool for tracking the etymology of Aramaic terms in the Talmud, or a calendar to track your daily Rambam studies. The AI assistant takes it from there, generating code to bring your vision to life.

With so much potential —and such low barriers to entry — it’s no surprise that this technology is being harnessed for Jewish learning.

Many people are taking advantage of this newly opened door to tech development. According to a recent article on Hashnode.com, 92% of US-based software engineers use AI tools daily, and a report from GitHub earlier this summer said a whopping 46% of new code is generated by an AI agent or assistant. Those numbers can be expected to climb as AI technology advances.

Today, anyone with an idea, a device, and an internet connection can bring a workable tool to life in an afternoon. And that trend is reflected in utilization of Sefaria’s data and API — digital building blocks that we make freely available for anyone looking to build their own Torah technology.

Over the past few months, the number of new Powered by Sefaria projects has been rising nearly daily. In June, we hit the milestone of 100 known projects built on Sefaria’s open data. As of early October, that number has climbed past 150. (In contrast, over the previous two years, we received notifications of less than 15 projects in total.)

It’s not just the number of projects, though — it’s also their variety. Increasingly, we see a wide variety of projects built to serve the specific needs of specific learners.

For example, Tehillim Reader was vibe-coded by someone who wanted Israeli children to be able to read Psalms during a period of hostility with Iran. Chevruta.AI, a source sheet builder, was created by a rabbi who wanted to help people assemble engaging learning materials. Lishkod is a free, Hebrew-first app that consolidates daily Torah study cycles into a single audio-enabled card feed with offline support and progress tracking.

Adi Dror, founder of Trade Mamba LLC, built Rambam Chat during the first Israel-Iran war. “At one point, I was sitting in the public bomb shelter in the middle of the night waiting for the next siren, so naturally I started playing around with Sefaria’s API and LLMs,” he says.

Even though he was in a shelter, Dror was able to use a combination of Sefaria’s data and AI tools, such as Ollama and models from Hugging Face, to build a functional chatbot that answers halakhic questions based solely on the Mishneh Torah (a seminal work by the 12th-century scholar Rambam).

“What made Sefaria useful here was simple: The text was already structured and accessible through the API, so I could spend my time on retrieval, embeddings, model behavior, and the chatbot itself instead of first building the entire data layer from scratch,” Dror says.

The combination of that existing data layer with an AI tool, however, meant he could “[take] what otherwise would have just been a notebook experiment and turn it into an actual chatbot with a simple web interface that I could send to someone over WhatsApp.”

In one night, a source-based, reliable chatbot was up and running.

The current uptick in vibe-coded projects is driven by AI coding assistants like Claude Code and Codex. But the reason Torah technologies specifically can flourish the way they have is Sefaria’s open-source data, API reference, and MCPs, all of which enable Jewish texts to be properly utilized by AI assistants.

Our code repository has always been open and free to use, an expression of our belief that Torah is the collective inheritance of the Jewish people. But making our data as accessible as possible goes beyond just sharing the digitized texts — it also means we can power a far broader ecosystem of Torah-based technologies than anything we could feasibly build on our own.

"All of Sefaria's data is open, which has always made it theoretically possible to build custom Torah learning experiences,” says Avi Levin, a software engineer at Microsoft and frequent collaborator with Sefaria’s engineers. “The recently released API documentation made it possible for average developers to think about building such an app, and obviously, agents that can read that documentation make it easier to take ideas and turn them into working apps. But one thing unappreciated about coding agents is how they can be pointed at Sefaria's open-source code to figure out how to handle texts, not just what the texts are."

To Levin, this distinction is why non-developers can now independently build Torah tools that actually work.

The power of AI, coupled with Sefaria’s robust codebase and API, is changing who gets to build Torah technology — and the results are already showing up across the Jewish internet.

Harrison Bleiberg teaches at the Alfred and Adele Davis Academy, a Jewish day school in Atlanta. Earlier this year, he told us about IvritSuite, his free collection of vibe-coded Hebrew learning tools for educators. "I am committed to making this a free hub of Hebrew learning tools," he wrote. "I see the impact that free, quality tools can have on a teacher's time and wellbeing."

Bleiberg credits Sefaria's open API and agentic coding platforms with making the suite, now visited thousands of times, possible. Sefaria's API powers IvritSuite's Torah Trainer, supplying the Hebrew text and translations for every parashah. The Trainer adds PocketTorah's open-source recordings of all 54 portions and haftarot, highlighting each word karaoke-style as it's chanted. Students can read in fonts from the Open Siddur Project, including scroll-like STaM scripts and dyslexia-friendly Hebrew, or in fonts they've designed themselves using IvritSuite's Font Maker.

Projects like Bleiberg’s that once required a developer are now being built by educators, students, parents, and learners who simply had a problem they wanted to solve.

There’s an added significance to what's happening here: People can now build Torah technology that meets a very specific learning need or interest. When a simple, workable app can be done in an afternoon with an AI assistant, mass appeal isn’t a prerequisite for a tool to be build-worthy.

“Sefaria’s goal has always been to open the doors of the metaphorical study hall so anyone anywhere can engage with Jewish texts — and that’s not theoretical,” says Sarah Kreitman, a Sefaria software engineer. “There are real learners, and I know many of them, whose schools or synagogues don’t have hard copies of canonical texts. For years, we’ve had the honor of opening the door to Torah learning regardless of physical access to texts. Now, with the combination of vibe-coding AI agents and Sefaria’s accessible, high-quality data, that ability to make Jewish texts available to all is super-charged.”

According to Kreitman, this combination means there can be as many avenues to Torah as there are learners. Niche learning needs or interests can be met with uniquely coded apps.

“I knew a woman once whose minimal Jewish education made Torah overwhelming to her — she didn’t know where to begin, even with prompts and suggestions,” Kreitman recalls. “Today, I could vibe-code her an app tailored to her specific goals and background in just a few minutes. That woman, today, could develop an independent relationship with the text, learning it in the way most compelling and suited to her.”

More access means more possibilities. We’re already seeing the impact vibe-coding is having on the Jewish learning experience. It’s the dyslexic students who can now study the weekly Torah portion, thanks to IvritSuite. It’s the grieving person who can find comfort in reading Psalms alongside commentaries, thanks to P'tiha B'Tehilim. It’s the curious learner who wonders about Kabbalah and Chasidut and can explore their intersection, thanks to AISefarim.com.

Each of these was vibe-coded. Each has hundreds of users — so far. Each serves a purpose otherwise unachievable.