Sitemap

Start here

• Home — what the lab is, and a guide to every section.
• Tools — the four things you can take away and use today.
• Solutions — finished products we have shipped, led by ScoobyBench.
• Dictionary — a 39-entry searchable reference for AI, mathematics, and software engineering.

Tools

• Convert — convert PNG, JPEG, WebP and AVIF in the browser. Nothing is uploaded.
• Converters — forty-odd conversions that run in the tab: CSV, JSON, Base64, hex, units, number bases, colour spaces, timestamps and image scaling.
• Extract — Turns PDFs, Word files, slides, spreadsheets and e-books into Markdown that keeps its headings and page boundaries, so the result can be split structurally rather than blindly.
• Chunk — Splits a Markdown document into retrieval-sized chunks along its own heading structure rather than at arbitrary offsets.
• Tokenize — Counts tokens and reports the distribution, including how many pieces exceed the model's context window and would be silently truncated at embed time.
• Embed — Turns chunks into vectors locally with BGE-M3 at 1024 dimensions or all-MiniLM-L6-v2 at 384.
• Index — Writes vectors and their chunks into a portable database directory you own, as Chroma or as plain numpy.
• Exploratory data analysis — Profiles a CSV from the command line and hands back the report, the figures, a replayable recipe and the script that reproduces them.
• Assistant — say what you want in plain English and it runs the tool here in the page.
• Prompts — eight prompts for real engineering work, free to copy.
• LoRA adapters — small adapters that stack onto a base model you already run.

Dictionary

• The Dictionary — 39 definitions anchored to papers and standard works across AI, mathematics, and software engineering.

Browse all 39 terms

• Acid
• Activation Function
• Attention Mechanism
• Backpropagation
• Bayes Theorem
• Bias Variance Tradeoff
• Big O Notation
• Cap Theorem
• Cohesion
• Convolutional Neural Network
• Coupling
• Cross Validation
• Deadlock
• Dependency Injection
• Eigenvector
• Embedding
• Entropy
• Eventual Consistency
• Gradient Descent
• Hash Table
• Idempotence
• Learning Rate
• Loss Function
• Markov Chain
• Maximum Likelihood Estimation
• Memoization
• Mutex
• Overfitting
• Principal Component Analysis
• Pure Function
• Race Condition
• Refactoring
• Regularization
• Reinforcement Learning
• Softmax
• Sql Injection
• Technical Debt
• Transformer
• Yak Shaving

Media

• Media — the entry point to everything we write and build.
• Blogs — a note board in two sections, filed by tag.
• Artifacts — interactive explainers you can play with.
• Time & Space Complexity — Big-O through the Master Theorem, with a live growth explorer.
• The Model Is No Longer the Product — prompting, RAG and finetuning, with playable demos.
• The Future of Local-First AI — why running models on your own machine is more than a privacy win.
• Raw: tokenisation, the questions — notebook page with an overview and a reading list.
• Raw: eight questions about LLM systems — notebook page with an overview and a reading list.

For machines

• sitemap.xml — the XML sitemap search engines read.
• llms.txt — a concise summary of this site for LLM crawlers.
• llms-full.txt — the full machine-readable directory, page by page.
• robots.txt — crawl permissions. Every AI agent is allowed.