AI for Data

AI Ad Manager

Generate ad copy variants at scale, brainstorm targeting angles, and use AI to analyze ad performance data for Google and Meta campaigns.

AI YouTube Management

Script videos faster, write titles and thumbnails that actually get clicked, and use AI to plan a channel content calendar and read your analytics.

Clean a Messy Dataset With AI

Take a genuinely messy spreadsheet and build a documented cleaning process. The rule throughout: AI proposes the rules, you apply them, and every change is reversible and logged.

Turn a Spreadsheet Into a Dashboard

Take data you already have and build a dashboard that answers a specific question. The discipline is deciding the question first, because a dashboard without one is just charts.

Set Up a CRM for a Small Business

Build a working customer record system for a real or imagined small business. The exercise is deciding what to track, which is where most CRM projects fail long before any software is involved.

The Wrong Chart Challenge

Deliberately visualise the same data four wrong ways to learn why the right one is right. Faster and more memorable than reading chart-selection guidance.

What Is RAG? Grounding AI in Your Own Documents

Retrieval-Augmented Generation lets a general model answer questions about your specific data without retraining. Here is how the pipeline works and the details that determine whether it works well.

Tokenizer

A tokenizer is the component that splits text into tokens before a model processes it. Different model families use different tokenizers, which is why the same text can have different token counts on GPT versus Claude.

Zero-shot prompting

Zero-shot prompting means asking a model to do a task with no worked examples — just an instruction. It works well for common tasks the model has seen extensively in training.

Structured output

Structured output means constraining a model to return data in a machine-readable shape such as JSON, usually by specifying the exact schema in the prompt or using a provider feature that enforces it.

Function calling

Function calling lets a model request that your code run a specific function with specific arguments. You expose tool definitions; the model returns a structured call, your code executes it, and the result goes back into the conversation.

Transformer

The transformer is the neural network architecture behind virtually all modern language models. Its key innovation is self-attention, which lets every token directly consider every other token in the input.

Parameters

Parameters are the learned numerical weights inside a model, adjusted during training. Parameter count — 7B, 70B, 405B — is a rough proxy for capacity, but not a reliable predictor of quality.

Multimodal model

A multimodal model accepts or produces more than one type of data — commonly text plus images, and increasingly audio and video. It processes them in a shared representation rather than through separate pipelines.

Open-weights model

An open-weights model has publicly downloadable parameters, letting anyone run it on their own hardware. Llama, Mistral, Qwen and DeepSeek are prominent examples.

Fine-tuning

Fine-tuning continues training a pre-trained model on your own examples to specialise its behaviour. It is the right tool for teaching consistent style, format or classification behaviour — not for adding knowledge.

Distillation

Distillation trains a smaller model to imitate a larger one, transferring much of the capability at a fraction of the inference cost. Most small production models are distilled from bigger teachers.

RAG (Retrieval-Augmented Generation)

RAG retrieves relevant passages from your own documents and inserts them into the prompt before the model answers. It grounds responses in your data, cuts hallucination, and needs no retraining.

Embedding

An embedding is a list of numbers representing the meaning of a piece of text, such that semantically similar texts have mathematically similar vectors. Embeddings make meaning-based search possible.

Vector database

A vector database stores embeddings and finds the most similar ones to a query vector quickly. It is the retrieval layer of most RAG systems. Examples include Pinecone, Weaviate, Qdrant and pgvector.

Chunking

Chunking splits documents into smaller passages before embedding them for retrieval. Chunk size is a key quality lever: too large dilutes relevance, too small loses the context needed to make sense.

Semantic search

Semantic search finds results by meaning rather than exact keywords, using embeddings to compare concepts. It matches "how do I get my money back" to a document titled "Refund Policy".

Reranking

Reranking takes an initial set of retrieved candidates and reorders them with a more accurate but slower model. It is one of the cheapest ways to materially improve RAG quality.

Knowledge cutoff

A model's knowledge cutoff is the date after which it has no training data. It cannot know about events, releases or prices after that point unless given the information in the prompt or via search.

Grounding

Grounding means tying model output to verifiable sources — retrieved documents, search results, or database records — rather than relying on parametric memory. Grounded answers can be checked.

MCP (Model Context Protocol)

MCP is an open standard for connecting AI models to external tools and data sources. It replaces bespoke per-integration code with one protocol any compatible client can use.

Tool use

Tool use is a model invoking external capabilities — search, code execution, database queries, APIs — to do things it cannot do from parameters alone, such as fetching current data or performing exact arithmetic.

Human in the loop

Human in the loop means inserting a person at decision points in an automated workflow — typically to approve consequential actions or review low-confidence outputs before they take effect.

Prompt injection

Prompt injection is an attack where malicious instructions hidden in content the model processes override the developer's intended behaviour. It is the most serious unsolved security issue in LLM applications.

Bias

Bias in AI is systematic skew in outputs that reflects patterns in training data or training process — including demographic stereotypes, cultural assumptions, and over-representation of dominant viewpoints.

JSON Schema

JSON Schema is a standard vocabulary for describing the shape of JSON data — its keys, types and required fields. AI providers use it to constrain model output to a guaranteed structure.

Cosine similarity

Cosine similarity measures how closely two vectors point in the same direction, on a scale from -1 to 1. It is the standard way to compare embeddings, because it captures semantic similarity while ignoring text length.

Weekly Automation Digest

Automated weekly team update via n8n + Slack

Structured Output (JSON & Tables) — ChatGPT

When you need output your app or spreadsheet can actually use, explicitly request a schema.

Structured Output (JSON & Tables) — Hugging Face (open-source)

When you need output your app or spreadsheet can actually use, explicitly request a schema.

Full-Funnel Message Mapping

Different funnel stages need different messages: problem-aware content at top, comparison and proof in the middle, urgency and risk-reversal at the bottom.

Reading Ad Performance Data with AI

Paste raw campaign metrics and ask AI to find the story in the numbers — which angle, audience, or creative is actually winning, and what to do next..

Original Insight: The Thing AI Can't Fake

AI averages the internet; audiences reward what the average doesn't contain — your data, your contrarian take, your first-hand story.

Reading Your Analytics with AI

Paste your YouTube Studio export and ask AI to find patterns across your best and worst performing videos — topic, length, hook style, and thumbnail pattern..

Series & Format Design

Channels grow on repeatable formats, not one-off videos: a recognizable structure (same open, same segments, same visual grammar) lowers production cost and trains the audience to return.

Project: One Video, Fully Engineered

Deliverable: a complete production package for one real video on your (or an invented) channel. Steps: 1.

Getting Cited by ChatGPT & Perplexity — Perplexity

AI answer engines favor content that is clearly structured, directly answers questions, cites data, and demonstrates first-hand expertise.

Getting Cited by ChatGPT & Perplexity — Claude

AI answer engines favor content that is clearly structured, directly answers questions, cites data, and demonstrates first-hand expertise.