Hundreds of thousands of Jews learn a page of Talmud a day. Most forget nearly all of it. This is a tool to change that β by teaching the one interpretive skill that recurs at every stratum of the halakhic tradition, and surfacing it, day after day, until it sticks.
The Daf Yomi cycle is arguably the most successful mass-participation learning program in Jewish history. A single page of Talmud, every day, for seven and a half years. Hundreds of thousands of participants worldwide. A new cycle begins roughly every eight years, each one larger than the last.
And yet β if you ask a committed Daf Yomi learner three years into the cycle what was on the page they studied last month, most cannot tell you. Ask them to reconstruct the core dispute of a tractate they "finished" two years ago, and the blank is honest and widespread. The ritual of daily learning is preserved. The learning itself is evaporating.
The existing publishers β Koren/Steinsaltz, ArtScroll, Mesivta, Hadran β are excellent at what they do. But what they do is help the learner understand today's daf. None of them are built to help the learner remember next month what they read today. The category doesn't exist.
The onboarding problem is even worse. A Jew who is drawn to Daf Yomi β in the postβOctober 7 Torah renaissance, this is a growing population β faces an impossible starting line. Joining mid-cycle means showing up to a page that references hundreds of prerequisite concepts, acronyms, and disputes the learner has never encountered. Most quietly drop out within weeks. The cost isn't just to them; every dropout is a participant the cycle never recaptures.
Every meaningful dispute in 2,000 years of Jewish legal literature has the same shape: a set of contested premises, a set of opinions each built from a specific combination of premise-stances, and a derivation that links foundation commitments to practical conclusion. That grammar is not specific to the Gemara. It recurs across strata:
This is not a metaphor. It is a structural feature of the literature. Teach a learner the grammar once, and they have the tool to read every layer. Surface it repeatedly β across different sugyot, different machlokot, different centuries β and each encounter reinforces the concepts they've already seen. Retention compounds because the atoms are reused.
The strata above do not form a one-way timeline from past to present. They form a graph. Each layer uncovers the layer immediately above it and immediately below it, in the same grammar. A learner can enter at any point and walk in either direction β from the pesukim in the Torah all the way down to the halacha as it is lived today, or from a ruling a person just received from their rav all the way back up to the Biblical verse at its root. Or stop anywhere in between. The graph is the same. The grammar is the same. Only the direction of travel is chosen by the learner.
The observant Jew with a practical question about Shabbat or kashrut. Starts at the ruling they received, walks it backward: the Shulchan Aruch that codified it, the Rishonim whose dispute the Shulchan Aruch resolved, the sugya in the Gemara the Rishonim were reading, and ultimately the pasuk in the Torah that anchors it all. They may stop at any step β by then the halacha is no longer a rule they keep, but a structure they understand.
The Daf Yomi learner who has just finished today's page. Wants to know why this argument matters beyond the academic. Walks it forward: the Rishonim who systematized it, the codifiers who resolved it, the contemporary posek whose ruling governs their home today. They may stop at any step β by then the abstract sugya is no longer a distant machloket, but the rule they will live with tonight.
This is the feature that serves both audiences β the observant learner asking "where does this come from?" and the Daf Yomi learner asking "why should I care?" Both questions, answered by the same traversal of the same graph. The grammar holds in both directions. That is what makes it a tool and not a textbook.
Click the Sugya Demo Β· English tab at the top of this page. You will find a working analysis of Hullin 2a β the opening page of the tractate that most Daf Yomi cohorts begin with, and one most learners find intimidating because it is filled with categories (Kuthi, Mumar, Meshumad) that feel arbitrary and archaic.
The demo re-presents that sugya in eight linked layers, each using the same visual vocabulary of color-coded premise pills:
A Hebrew version is one tab over β identical in content, mirrored for RTL reading, so the tool is accessible to learners in either language from day one.
The demo is a proof of concept, not the product. It answers one question: does the grammar actually fit a real sugya at depth? It does. The work now is to build the infrastructure that lets this depth be produced at the pace of the cycle β one daf per day β and surfaced to learners in a form that drives retention.
The existing publishers have solved the problem of understanding today's daf. They have done it beautifully, in print and digitally, for generations. None of them are going to be displaced, and this project does not attempt to displace them. A serious Daf Yomi learner will continue to read Steinsaltz or Koren β as they should.
The problem this project addresses is the one none of them are designed to solve: retention across the cycle, onboarding for newcomers, and structural coherence across strata.
| Existing publishers | This tool | |
|---|---|---|
| Primary goal | Understand today's daf | Retain the daf across the cycle |
| Unit of content | A line of text, a note, a summary | A recurring concept in a cross-sugya graph |
| Learner interaction | Read | Predict, generate, re-derive, recall |
| Progress tracking | None | Per-concept mastery, spaced review |
| Onboarding for newcomers | "Start reading" | "Here are the 20 prerequisites for today's daf β learn those first" |
| Handling of pluralism | Editorial (usually a single tradition) | Structural (premises, opinions, dissent preserved) |
| Layers of tradition | Typically one (Gemara + Rashi) | All four, same grammar (Gemara β Rishonim β Shulchan Aruch β contemporary) |
The relationship is complementary. This project will link out to the primary-text publishers rather than reproduce them. Its value proposition is distinct: not the text, but what the learner retains of the text three years after encountering it.
The current (14th) Daf Yomi cycle ends in June 2027. A new cycle begins the next day, drawing in the largest cohort of beginners the program sees every seven and a half years. A tool launched in sync with that moment reaches the widest possible audience at exactly the moment they most need it. A tool launched six months after misses the onboarding window entirely and has to wait until 2034 for the next one. That timing is not a convenience β it is the central strategic constraint of the project.
The implication: we need to be hardened and ready by January 2027 for a soft launch, and fully polished by June 2027 for the cycle kickoff. That gives roughly a year of focused work.
It is also a year we need, for a reason beyond software. Authoring has lead time. Every sugya in the cycle has to be drafted, reviewed rabbinically, proofread, and structurally tested before a learner ever sees it. Even with AI-assisted drafting, a human editorial review cannot be compressed below a certain floor. The authoring pipeline has to start now to build a reviewed backlog of sugyot ready for daily release when the cycle begins. Waiting until launch to start authoring guarantees falling behind the cycle in the first month β which is how every prior attempt at a daf-yomi-synced product has failed.
Two additional forces make the moment especially favorable. The authoring bottleneck has collapsed: a depth of analysis that would have taken a team of scholars months per sugya a decade ago can now be drafted in hours with AI under editorial control. And the audience has expanded: in the postβOctober 7 Torah renaissance, a documented surge of formerly unaffiliated Jews has turned toward serious learning β precisely the population locked out by existing tools.
The work is staged so each stage validates the one before it. Nothing here requires a leap of faith, and the order is dictated by the June 2027 launch constraint.
One sugya (Hullin 2a) built end-to-end at eight layers of depth, bilingual, interactive. Visible in the Demo tabs. Validates that the structural grammar fits a real Gemara at full depth, and that a learner can re-derive the whole sugya from four questions once the foundations are laid.
The strongest version of this project is funded the way Torah projects have always been funded: by combining a leadership gift that capitalizes the build with a community of dedicators who sustain it cycle after cycle. Both pieces matter. Neither works alone.
One transformational gift that capitalizes the build phase: editorial team, rabbinic review, engineering, infrastructure, and the authoring backlog needed to launch with the cycle in June 2027.
In recognition: Founding Patron of the project in perpetuity. The Patron's name appears on the project's home page and on every dedicated daf alongside the daily sponsorships, identifying the person without whom none of the rest exists.
Without this gift, the sponsorship engine has nothing to sponsor.
Once the system exists, every daf can be dedicated by a family, individual, or institution β l'iluy nishmat, in honor of a simcha, on a yahrzeit, or as a recurring annual dedication. Every learner who opens that daf sees the dedication.
This is the model that funds ongoing authoring, review, operations, and team costs β not for a year, but for every Daf Yomi cycle that follows. Revenue scales with adoption.
A self-sustaining system in the tradition of Torah philanthropy.
Sponsorship is structured as a familiar ladder, anchored at chai-multiples that match the conventions of Torah philanthropy. A sponsor can dedicate a single daf for a yahrzeit, a perek for a milestone, a full masechet for a parent, or β at the prestige tier β the entire Shas in one transformational gift.
A small banner appears at the top of the page, in the sponsor's chosen wording. Tasteful, prominent, never intrusive. Below is a sample of what a learner opening the page would see:
Sample dedication. Wording is fully customizable: l'iluy nishmat, l'zecher, in honor of a wedding, bar/bat mitzvah, refuah sheleima, or simply with gratitude.
David Schwartz has been learning Daf Yomi for more than four decades, having joined his first cycle at the age of thirteen and currently working through his sixth. He has given shiurim β public Talmud classes β for the majority of his adult life, and is now studying for semicha (rabbinic ordination), which has brought the Shulchan Aruch and the contemporary halakhic stack into focus alongside the daily Gemara.
Professionally, David is an entrepreneur who has spent the last several years working intensively with AI systems. The intersection of those two lives β the learner who has watched generations of fellow daf yomi participants struggle with retention, and the builder who has seen firsthand how rapidly AI is reshaping what educational tooling can do β is what makes this project feel both necessary and newly feasible. AI cannot replace the chavruta, the rav, or the text. But used with discipline and under rabbinic review, it can do something the existing tools were never designed to do: lower the barrier for newcomers, structurally represent the pluralism of the tradition, and help the daily learner actually retain what they are reading.
This project is the place where those three strands β the learner of forty-four years, the teacher preparing for semicha, and the technologist β converge.
Next step: a 30-minute conversation to walk through the demo live, discuss where the thesis feels strong and where it feels underspecified, and align on what success looks like in time for the June 2027 launch.
Contact: david@chavruta.ai