Projects
Two tools built at the intersection of Buddhist practice and AI — a public dharma reference library for practitioners, and a private research workbench for teachers. Both are citation-grounded, non-commercial, and built from inside the Insight/Vipassanā tradition.
Kalyāṇa Mitta
"Admirable friendship, admirable companionship, admirable camaraderie — this is the whole of the holy life." — SN 45.2, Upaḍḍha Sutta
A citation-grounded dharma reference tool for practitioners in the Insight/Vipassanā tradition. Ask a question — "What does the Buddha teach about grief?" or "How do teachers talk about belonging?" — and receive passages from the Pāli canon alongside how contemporary Insight teachers have engaged the same teaching.
It is a library, not a teacher. Every passage is attributed, linked to its source, and offered as a bridge to a living teacher and sangha — never as a substitute. The corpus spans 116 suttas, the full Dhammapada, the Therīgāthā, and 34 teachers from Mahasi Sayadaw and Ajahn Chah through Ruth King, Larry Yang, Zenju Earthlyn Manuel, and Lama Rod Owens.
Corpus: SuttaCentral (CC0) · Inquiring Mind archive (CC BY-NC-ND 4.0) · IMC Written Dharma
Domain: dharmadesk.org (forthcoming)
Sutta Desk
"The whole of the dharma is available to you in this body, in this mind, in this moment." — Anagarika Munindra
A private research workbench built on the same corpus as Kalyāṇa Mitta, designed for teachers doing talk preparation. Ask a research question and receive canonical ground, lineage transmission, contemporary teacher voices, teaching angles, tensions and cautions, and an honest account of what the corpus is missing.
For teachers doing the work of preparation — not a replacement for sitting with the texts, but a research partner that knows the library. The three-tier structure (canon → lineage transmission → contemporary teachers) makes the transmission chain visible in every response.
Guest access: Email wayneataylor@gmail.com to request the passphrase
Questions and answers
What is this, exactly?
A retrieval tool, not a teacher. When you ask a question, the tool searches a fixed library of canonical texts and teacher writings, finds the most relevant passages, and presents them with full attribution. It does not generate dharma teachings, paraphrase the Buddha's words, or offer its own spiritual guidance.
Where do the canonical passages come from?
All sutta translations are by Bhikkhu Sujato, sourced from SuttaCentral and released under CC0 (public domain). Passages are quoted directly — never altered, paraphrased, or summarized — and every response links back to the original on SuttaCentral.
Where do the teacher passages come from?
Curated quotations from published books, drawn from teachers across the Insight/Vipassanā lineage — from Mahasi Sayadaw and Ajahn Chah through the Western Insight generation to contemporary BIPOC voices. Every passage is attributed with author, title, publisher, and year. Quotes are verified against original source text wherever possible.
What's the "Source Guardian" process?
A commitment to seek explicit permission before including any teacher's recorded talks or an organization's archived content — not just relying on what's technically accessible online. Outreach to AudioDharma and Dharma Seed is underway; their content is not included until permission is granted.
Does the AI ever make things up?
The tool is instructed to only draw from the retrieved passages and to say plainly when the corpus doesn't contain a good answer to a question — rather than inventing a citation or improvising a teaching. If something seems off, that's worth flagging — see contact info below.
What's actually in the corpus right now?
116 canonical texts and 34 teachers as of this writing, spanning the full Dhammapada, the Therīgāthā, the aggregates cycle, the brahmavihārās, the five spiritual faculties, the seven awakening factors, jhāna and samādhi, core CDL curriculum texts, and contemporary BIPOC dharma voices. Thanissaro Bhikkhu's writings (CC BY-NC 4.0) are also included as a supplementary source. The corpus grows steadily — see the full list below.
Full corpus list →On ethics and AI
These tools are built with care for the tradition. Neither tool generates dharma teachings — they retrieve and attribute source passages. All canonical texts are drawn from openly licensed translations. Teacher passages are sourced from published books with full attribution, and teacher permissions are being sought through a Source Guardian process modeled on the tradition's commitment to honest transmission.
The tools are non-commercial, offered freely, and designed to direct practitioners toward living teachers and sanghas — not to replace them. For questions about corpus inclusion, attribution, or the Source Guardian process, contact wayneataylor@gmail.com.
Environmental impact and AI
These tools are not carbon-neutral — no AI tool is. But I built them with efficiency in mind: small Anthropic models where possible, hosting on renewable-energy infrastructure (AWS Oregon, Vercel), on-demand processing rather than running continuously, and a manually refreshed corpus that minimizes unnecessary compute. Per query, the energy footprint is comparable to a web search.