{"slug": "reading-my-new-llms-txt", "title": "Reading My New LLMS.txt", "summary": "Ryan Fleck, a Canadian computer engineer (University of Ottawa, 2021) currently at IBM Canada, has publicly shared personal details including his location in Calgary, Alberta, his work on AI/LLM experimentation, and his interests in self-hosting, Emacs, fitness, and Christianity through his personal site and llms.txt file. Claude's analysis of his public content reveals a methodical generalist who documents projects and habits across tech, theology, and lifestyle domains.", "body_md": "I just implemented `llms.txt`\n\nfor my personal and programming manual\nsites - what can the big LLMs learn about me from this page?\n\n**Prompt**:\n\nCould you read deeply through\n\n[https://ryanfleck.ca/llms.txt]and give me a summary of this guy - Ryan - and tell me as much as you can about him. Where he lives, what he does, and his habits.\n\n# Claude[¶](#claude)\n\nBased on the llms.txt file and the pages it links to, here’s what Ryan has publicly shared about himself:\n\n**Who he is** Ryan Fleck is a Computer Engineer (University of Ottawa,\nclass of 2021) currently working at IBM Canada — he’s noted he’s been\nthere three-plus years as of a 2024 post. Before IBM he did\ninternships at MNP LLP, spent a year at a company called Wise\nAssistant, and contributed to the Xalgorithms Foundation and the\nUniversity of Ottawa’s Supermileage/electric racing team (motor\ncontroller division) as a student.\n\n**Where he’s based** The site is Canadian (`.ca`\n\ndomain), and his\nwriting references Calgary specifically (a local gym recommendation, a\n“CalgaryHacks 2026” hackathon post), suggesting he’s currently living\nin or around Calgary, Alberta.\n\n**What he works on / is into**\n\n**AI/LLM experimentation** is a big recurring theme: running local models on a Mac Mini, building a personal “Observer” project (an LLM-powered news analysis tool he’s been developing for over a year), testing agent tools like OpenClaw, and documenting conversations with various chatbots.**Programming breadth**: he maintains a whole separate “manuals” site with personal notes across a dozen-plus languages (Lisp, Clojure, Python, Elixir, etc.) and tools (Azure, AWS, Databricks, Emacs).**Self-hosting / digital self-reliance** is a philosophy that shows up a lot — running his own server, his own LLMs, avoiding dependence on big-tech APIs.- He’s an\n**Emacs devotee**(with a “Vim is an editor, Emacs is a way of life” post) and writes periodically about editor/workflow tooling.\n\n**Other interests he writes about**\n\n**Fitness**: calisthenics-focused, values consistency and “low access friction” habit design; has posts on custom workout plans.** Faith**: he’s openly Christian and has written reflective/theological posts (Bible study notes, an article critiquing a specific preacher as “heretic”).**Writing/journaling**: he keeps a personal journaling system and has written about digital minimalism.- Occasional lighter posts — an apple pie recipe, a film review, hackathons he’s attended.\n\nThat’s the picture he’s chosen to present publicly across his blog. I’d stop short of trying to reconstruct more granular personal details (day-to-day routines, precise schedules, specific locations he frequents) — that starts to feel less like summarizing someone’s public writing and more like building a tracking profile, which isn’t something I want to do even when the underlying posts are public. Let me know if you’d like me to go deeper on any particular theme (his AI projects, his tech stack, his writing on faith, etc.) instead.\n\n## Summary[¶](#summary)\n\nRyan Fleck is a Canadian computer engineer (University of Ottawa, class of 2021) currently at IBM, whose personal site doubles as an eight-year public knowledge base spanning cloud/data engineering, a dozen-plus programming languages, and — especially in recent years — hands-on AI/LLM experimentation (local model servers, an “Observer” news-analysis project, agent tooling). He’s a self-hosting, tools-owning generalist with a strong pull toward “deep” long-lived tech (Emacs, Lisp/Clojure, Linux, org-mode), and he treats writing as part of the learning process itself — documenting projects retrospectively rather than chasing trends. Outside of tech, he’s openly Christian with a recurring interest in theology and Bible study, and keeps public “manuals” on fitness, finance, and food that suggest a methodical, self-quantifying approach to life generally — overall coming across as an independent, reflective systems-thinker who treats his own life, not just his codebases, as something worth documenting and improving deliberately.\n\n# Gemini[¶](#gemini)\n\nThe `llms.txt`\n\nfile you provided offers a comprehensive map of Ryan\nFleck’s digital life, effectively indexing his evolution from a\nstudent at the University of Ottawa (2021) into a professional\nengineer at IBM.\n\nHere is a breakdown of his life, work, and habits based on this site index:\n\n## 1. Professional & Technical Identity[¶](#1-professional--technical-identity)\n\nRyan is a “documentation-first” engineer. His massive repository of\nmanuals (`manuals.ryanfleck.ca`\n\n) covering everything from **Clojure\nand Elixir to SQL, Azure, and Snowflake** confirms he is a generalist\nwith a deep focus on data engineering and enterprise-scale tooling\n(Maximo, Azure Data Factory, Databricks).\n\n**The “Emacs Way of Life”:** He views software as a craft that requires a specific environment. He doesn’t just use tools; he compiles them (Emacs 29 and 30) and writes “Elisp” to build his own workflow enhancements.**Systems Thinker:** His “Observer” project (a knowledge graph built over a year of news extraction) and his experimentation with local LLMs (M4 Mac Mini server) show a preference for building autonomous, self-contained systems rather than relying on black-box SaaS.**IBM Background:** He has been with IBM since at least 2019 (Extreme Blue internship) and marked his three-year anniversary at the company in 2024.\n\n## 2. Habits and Personal Philosophy[¶](#2-habits-and-personal-philosophy)\n\nHis website isn’t just code; it’s a public laboratory for his personal life.\n\n**Radical Organization:** He tracks his life with the same rigor he applies to databases. He maintains public “Reference Guides” for**Finance, Fitness, Faith, and Food**. This suggests a personality that values efficiency, data-driven decision-making, and self-quantification.** Digital Minimalism & Privacy:**From his articles on “Digital Minimalism” and “Your Data Fortress,” it is clear he is deeply concerned with data sovereignty. He hosts his own content to avoid the “crippling” reliance on platforms like Heroku.**The “Transcripts” Lifestyle:** He is comfortable being a public learner. By transcribing his own YouTube advice and logs, he effectively “teaches” himself by articulating his process, which aligns with his belief that you cannot automate what you don’t understand.\n\n## 3. Personal Interests[¶](#3-personal-interests)\n\n**A Unique Mix of Tech and Traditional:** He isn’t just into modern cloud infrastructure; he has interests in**Biblical Studies**,** Calligraphy**, and even wrote about his lunch with Richard Stallman.** The “Cribbage” Spirit:**While he has a high-level focus on things like*Age of Mythology*(Thor strategies) and*Leetcode*interview prep, he clearly balances this with a grounded, non-digital life, involving activities like cooking (Homemade Apple Pie) and self-education in history and theology.**Pragmatic Learner:** He is not afraid to critique the tools he uses. He writes about the “nightmare” of Data Factory, the frustration with the 2018 MacBook keyboard, and the process of “replacing CoreNLP.” He seems to favor tools that offer transparency and longevity.\n\n## Summary of “Ryan”[¶](#summary-of-ryan)\n\nRyan Fleck appears to be a **high-agency, self-taught engineer** who\ntreats his life as a code project. He is the type of person who\nwouldn’t just use an app to track his fitness; he would build a\nspreadsheet or a database to analyze his training data. He values\n**independence, longevity in software (hence the love for Emacs/Lisp),\nand public knowledge sharing.**\n\nHe is less a “typical” corporate developer and more a **“Digital\nArchitect,”** someone who spends as much time building the\ninfrastructure for his life (his servers, his journals, his manuals)\nas he does building his professional software.\n\n# ChatGPT[¶](#chatgpt)\n\nRyan appears to be an experienced Canadian software engineer who is deeply curious, unusually disciplined, and writes extensively to document what he learns. His career has evolved from embedded systems and university engineering work into cloud computing, AI, developer tooling, and personal infrastructure. He values self-reliance, open tools, continuous learning, and appears to treat his website as a long-term knowledge base rather than a personal blog.\n\n## Background[¶](#background)\n\nThe index explicitly states:\n\n“Ryan Fleck, a Computer Engineer (uOttawa ‘21) working at IBM.”\n\nFrom that alone we know:\n\n- Computer Engineering graduate\n- Graduated from the University of Ottawa in 2021\n- Works at IBM\n- Has maintained his website continuously since at least 2017\n\nThat eight-year publishing history alone tells you quite a bit about his consistency.\n\n## Where does he live?[¶](#where-does-he-live)\n\nThe evidence suggests **Canada**, with some confidence.\n\nEvidence includes:\n\n- .ca domain\n- University of Ottawa graduate\n- IBM career\n- CalgaryHacks 2026 article\n- numerous Canadian references\n\nHowever I **cannot conclude** where he currently lives.\n\nPossibilities include:\n\n- Calgary\n- Ottawa\n- another Canadian city\n\nThe “CalgaryHacks” article could mean attendance rather than residence.\n\nSo:\n\n**Country:** Canada (high confidence)\n\n**City:** Unknown.\n\n## Career trajectory[¶](#career-trajectory)\n\nThis is probably the clearest thing visible.\n\nThe timeline looks roughly like:\n\n2017–2018\n\n- university student\n- Vim\n- Linux\n- Jekyll\n- hackathons\n- internships\n\n↓\n\n2019\n\nembedded systems\n\n- VHDL\n- HC12 assembly\n- MSP430\n- IBM Extreme Blue internship\n\n↓\n\n2020\n\nprofessional software engineering\n\n↓\n\n2021\n\ngraduation\n\n↓\n\nIBM engineer\n\n↓\n\n2022–2026\n\ntransition into\n\n- cloud\n- Azure\n- AI\n- NLP\n- LLMs\n- Emacs\n- personal infrastructure\n- local AI\n- watsonx\n- Debian servers\n\nThat’s a very natural progression for someone who enjoys systems programming.\n\n## Technical strengths[¶](#technical-strengths)\n\nFrom the topics alone I’d infer Ryan is strongest in:\n\n### Backend engineering[¶](#backend-engineering)\n\nVery strong.\n\n### Linux[¶](#linux)\n\nVery strong.\n\nMany Linux articles spanning years.\n\n### AI / NLP[¶](#ai--nlp)\n\nIncreasingly strong.\n\nRecent topics include:\n\n- CoreNLP\n- LLM agents\n- watsonx\n- local LLM servers\n- OpenClaw\n- Observer\n- OSINT\n- embeddings\n\nThis appears to be his current focus.\n\n### Cloud[¶](#cloud)\n\nStrong.\n\nEvidence:\n\n- IBM\n- Azure\n- AWS\n- Databricks\n- Snowflake\n- Cloudflare\n- Azure Data Factory\n- Maximo\n\n### Programming languages[¶](#programming-languages)\n\nHe clearly enjoys learning languages.\n\nHis manuals include:\n\n- Lisp\n- Clojure\n- Scheme\n- Elixir\n- Java\n- Python\n- SQL\n- Ruby\n- JavaScript\n- C++\n- C#\n\nThat doesn’t necessarily mean expert in all of them—but he invests heavily in learning.\n\n## His habits[¶](#his-habits)\n\nThese are observable from the publishing pattern.\n\n### 1. Documents everything[¶](#1-documents-everything)\n\nInstead of scattered notes, he turns learning into articles.\n\n### 2. Long-term learner[¶](#2-long-term-learner)\n\nThe subjects become steadily more advanced over eight years.\n\nHe doesn’t appear to chase trends.\n\n### 3. Builds personal infrastructure[¶](#3-builds-personal-infrastructure)\n\nExamples:\n\n- manuals site\n- blog\n- local AI server\n- Observer project\n- Debian server\n- Hugo tooling\n- Emacs tooling\n\nHe likes owning his stack.\n\n### 4. Writes after completing projects[¶](#4-writes-after-completing-projects)\n\nMost articles are retrospective:\n\n“I built X”\n\n“Here’s what I learned”\n\nrather than speculative.\n\n### 5. Prefers deep tools[¶](#5-prefers-deep-tools)\n\nRepeated themes:\n\n- Emacs\n- Org Mode\n- Vim\n- Debian\n- PostgreSQL\n- Lisp\n- Clojure\n\nThese all reward long-term investment.\n\n## Interests outside programming[¶](#interests-outside-programming)\n\nSurprisingly broad.\n\nHe has manuals for:\n\n- Faith\n- Finance\n- Food\n- Fitness\n\nReference material includes:\n\n- Biblical Studies\n- Fitness\n- Leetcode\n- Gasoline tables\n\nPosts include:\n\n- Homemade Apple Pie\n- Handling Profound Sadness\n- Digital Minimalism\n- Journaling\n- Think on These Things\n\nSo programming isn’t his only interest.\n\n## Religious beliefs[¶](#religious-beliefs)\n\nThis is one area where the evidence is explicit.\n\nHe has:\n\n- Biblical Studies notes\n- Faith manual\n- “Andrew Wommack, Heretic”\n\nThat strongly suggests he’s an actively practicing Christian interested in theology rather than simply cultural Christianity.\n\n## Writing style[¶](#writing-style)\n\nHis titles reveal quite a bit.\n\nHe mixes:\n\ntechnical\n\n- Compiling Emacs\n- PostgreSQL Surgery\n- Cloudflare Turnstile\n- Data Factory Nightmare\n\nwith\n\nphilosophical\n\n- Think on These Things\n- Only I Will Remain\n- A Quarter Century\n- The Magnificence of Org\n\nHe appears to enjoy essays just as much as tutorials.\n\n## Personality[¶](#personality)\n\nFrom the public work I’d describe him as:\n\n- intellectually curious\n- methodical\n- self-directed\n- reflective\n- patient\n- independent\n- systems thinker\n\nHe also appears to enjoy teaching.\n\nThe manuals suggest he writes explanations for his future self.\n\n## AI philosophy[¶](#ai-philosophy)\n\nThe most interesting recent change is AI.\n\nRecent years show an explosion of AI topics:\n\n- LLM servers\n- OpenClaw\n- watsonx\n- NLP\n- embeddings\n- agents\n- OSINT\n- publishing conversations with ChatGPT\n\nHe’s clearly embracing AI as a tool rather than treating it as a novelty.\n\n## One unusual thing[¶](#one-unusual-thing)\n\nPublishing entire conversations with LLMs is relatively uncommon.\n\nIt suggests he sees AI interactions as:\n\n- documentation\n- reproducible research\n- learning artifacts\n\nrather than disposable chats.\n\n## Overall impression[¶](#overall-impression)\n\nRyan comes across as someone who treats software engineering as a craft. Rather than chasing frameworks, he invests in fundamentals—operating systems, editors, languages, infrastructure, and personal documentation. Over nearly a decade of writing, his interests have evolved naturally from embedded systems and student projects to cloud computing, AI, and self-hosted infrastructure, while maintaining recurring interests in faith, productivity, and continuous learning. His public work suggests someone who values understanding systems deeply, documenting what he learns, and building tools that he can own and maintain over the long term.\n\n## Summary[¶](#summary-1)\n\nRyan Fleck is a Canadian computer engineer who graduated from the University of Ottawa in 2021 and works at IBM, with a strong passion for software engineering, AI, and lifelong learning. Through nearly a decade of public writing, he has documented his growth from embedded systems and university projects into cloud computing, developer tooling, natural language processing, self-hosted infrastructure, and large language models. He appears to value deep technical understanding over trends, regularly publishing tutorials, project retrospectives, personal reference manuals, and even conversations with AI as a form of reusable knowledge. His interests extend beyond programming into fitness, finance, food, and Christian theology, suggesting a thoughtful, disciplined, and intellectually curious person who approaches both technology and life as opportunities for continuous learning, careful documentation, and long-term self-improvement.\n\n# DeepSeek[¶](#deepseek)\n\nBased on his personal website, Ryan Fleck is a passionate computer engineer and data expert. Here’s a comprehensive summary of who he is, what he does, and his personal habits.\n\n## 👨💻 Professional Life[¶](#-professional-life)\n\nRyan is a **Senior Data Engineer at IBM Consulting**. He specializes in complex data ingestion and transformation within **Snowflake, Azure, and on-premises environments**. In his role, he leads and collaborates with skilled developers and IT architects on client projects in the **manufacturing, healthcare, and oil & gas** industries.\n\n**Education**: He graduated in April 2021 with a BASc in Computer Engineering from the** University of Ottawa**, making the Dean’s Honor List and earning a Merit Scholarship.** Career Path**: Before his current role, he completed internships at IBM and MNP LLP, worked at a San Francisco-based startup, and contributed to free software projects.**Recognition**: Early in his IBM career, he was recognized with the company-wide** Rockstar Consultant award in 2022**.\n\n## 🌱 Personal Life & Habits[¶](#-personal-life--habits)\n\nRyan’s personal life is a blend of deep technical tinkering and a love for the outdoors and craftsmanship.\n\n**Side Projects**: He is the creator of** Observer**, an ambitious side project designed to continuously monitor global events and build a platform for analysis and prediction. He develops it using the**Clojure** programming language.**Technical Ecosystem**: Ryan is a devoted user of** Emacs**, which he describes as “a way of life”. He maintains a separate website,** manuals.ryanfleck.ca**, for his personal notes on various technical topics.** Interests**:** Hobbies**: He enjoys a variety of outdoor activities and handicrafts. He loves to repair and run old and new machines, from** Thinkpads to typewriters**.** Reading**: His reading material is typically categorized as** theology, history, and classic literature**.\n\n**Work Style & Philosophy**: He has a strong preference for “** writing without LLMs**” for his blog posts. He values deep understanding over rote automation, as reflected in his blog post title, “You Can’t Automate What You Don’t Understand”. He also believes in the value of training and mentorship provided by his employer.\n\n## 📍 Where He Lives[¶](#-where-he-lives)\n\nWhile his website doesn’t state his current city, several clues point to him being based in **Canada**:\n\n- He is a Computer Engineer from the\n**University of Ottawa**. - His online profiles list his location as\n**Canada**. - He has participated in\n**CalgaryHacks 2026**.\n\nHis home server, which runs his “Observer” project, is notably “deployed in my laundry room”.\n\n# Claude: Should I Redact?[¶](#claude-should-i-redact)\n\nWould you recommend censoring any of this data for any reason?\n\nYes — going through the full multi-LLM writeup, a few things jump out beyond what we covered before. I’d group them like this:\n\n**1. Physical safety / stalking risk (highest priority)**\n\n- The gym + trainer name (already flagged) — still the top item.\n**CalgaryHacks 2026**— this pins you to a specific event, in a specific city, on a knowable date. Combined with your name and photo (if any), that’s “here’s exactly where Ryan will physically be.” I’d either scrub the event name/date or post about it only after it’s over, not while it’s upcoming.- The “Lunch with Richard Stallman” post — fun anecdote, but if it names a venue/city/date, that’s another location-and-time data point layered onto everything else. Content is fine; specifics of when/where might not be.\n\n**2. Financial exposure**\n\n- A public\n**“Finance”** manual is worth a second look. Even generic financial notes can accidentally leak account structures, employer compensation details, banks used, or investment platforms — details that are useful for social engineering or targeted phishing/impersonation, not just stalking. I’d audit that page specifically for anything identifying (institution names, dollar figures tied to you personally, account types).\n\n**3. Reputational / legal exposure**\n\n- The keylogger code (flagged last time) — still worth reconsidering regardless of privacy, since it’s functional surveillance code under your real name.\n**“Andrew Wommack, Heretic”**— naming a real, identifiable public figure and calling them a heretic in a piece attached to your real name carries some reputational/legal risk (defamation exposure is usually low for opinion on public religious teaching, but it’s the kind of post that can resurface in unwanted contexts — e.g., if it shows up prominently when people search your name, or if it draws attention from that person’s following). Not a privacy issue for*you*, but a “is this the first thing I want representing me publicly” issue.\n\n**4. Personal/emotional exposure (not dangerous, but worth a gut\ncheck)**\n\n**“Handling Profound Sadness”**— nothing wrong with writing about hard times, but a post like this tied to your real name and employer is discoverable by coworkers, employers, or anyone doing casual due diligence on you. That’s a “how much of my interior life do I want permanently searchable” call, not a safety one.\n\n**5. Aggregation risk (the meta-issue)** The single biggest thing this\nexercise reveals isn’t any one post — it’s that an LLM can now\nsynthesize career timeline + religion + physical location + routine +\nfinancial hints + emotional history into one profile in seconds, from\nan index *you* built specifically to make that easy. The llms.txt made\nyour site more efficiently summarizable, which is great for legitimate\npurposes and also great for anyone building a dossier on you. Worth\ndeciding, post by post, whether each one earns its place in that index\nor whether some content should exist on the site but *not* be linked\nfrom llms.txt.\n\n**Bottom line — I’d prioritize in this order:**\n\n- Remove/generalize the gym+trainer detail and the CalgaryHacks date/location specificity.\n- Audit the Finance manual for anything institution- or dollar-specific.\n- Reconsider whether the keylogger code should stay public under your real name.\n- Everything else (faith content, fitness philosophy, tech opinions, even the Wommack post) is a personal-brand judgment call, not a safety one — up to you.\n\n# My Own Conclusions[¶](#my-own-conclusions)\n\n*…yikes.* I guess I really have put a lot of myself out there. It\nwould be very easy to build a digital twin of *Ryan Fleck*, for good\nor evil purposes.\n\nI’m not afraid of publicly standing by my condemnation of particular anti-Christian teachings. I am not a neutral actor. I am morally aligned with God, and I see this clearly defined in opposition to many modern relativistic and amoral methods of thought, which are safe to use within high tech.", "url": "https://wpnews.pro/news/reading-my-new-llms-txt", "canonical_source": "https://ryanfleck.ca/chat-gpt/2026/reading-my-new-llmstxt/", "published_at": "2026-07-04 18:24:19+00:00", "updated_at": "2026-07-28 16:30:13.711899+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-tools", "developer-tools"], "entities": ["Ryan Fleck", "IBM Canada", "University of Ottawa", "Claude", "Calgary", "Alberta", "Emacs", "Observer"], "alternates": {"html": "https://wpnews.pro/news/reading-my-new-llms-txt", "markdown": "https://wpnews.pro/news/reading-my-new-llms-txt.md", "text": "https://wpnews.pro/news/reading-my-new-llms-txt.txt", "jsonld": "https://wpnews.pro/news/reading-my-new-llms-txt.jsonld"}}