← All masterclasses

Generative AI for Analytics Teams

From prompts to pipelines — turn raw data into insights with LLMs and agents.

3–4 days Data analysts, analytics engineers, BI leads

Overview

Your analytics team spends too much time writing SQL, cleaning data, and formatting reports. This masterclass teaches them to use LLMs and agentic pipelines to automate the repetitive parts — from natural language to SQL, automated EDA, to self-updating dashboards — while keeping humans in the loop for decisions that matter.

What you’ll build

  • A natural-language-to-SQL interface for your data warehouse
  • An automated exploratory data analysis agent
  • A report generation pipeline that writes and formats insights in natural language
  • A conversational analytics assistant your team can query

Curriculum

Day 1 — LLM Foundations for Analytics

  • Prompt engineering for structured output (SQL, JSON, tables)
  • Text-to-SQL: approaches, limitations, and safety
  • Connecting LLMs to your data warehouse (Snowflake, BigQuery, Postgres)
  • Output validation: ensuring generated SQL is safe and correct
  • Hands-on: build a text-to-SQL interface on your schema

Day 2 — Automated EDA & Data Profiling

  • Using LLMs to generate statistical summaries and anomaly flags
  • Automated chart selection and generation (Matplotlib, Plotly, ECharts)
  • Data quality checks with LLM-assisted validation
  • Multi-dataset correlation discovery
  • Hands-on: build an automated EDA agent that profiles uploaded CSVs

Day 3 — Report Generation & Narratives

  • Template-based vs free-form report generation
  • Grounding insights in data: citation and traceability
  • Automated chart annotation and commentary
  • Scheduling and incremental report updates
  • Hands-on: build a weekly report generator for a real dataset

Day 4 — Production Pipelines & Governance

  • Orchestrating analytics agents with task queues
  • Caching strategies for expensive queries
  • Access control: ensuring the LLM only sees permitted data
  • Audit trails for AI-generated insights
  • Cost management: when to use large vs small models
  • Capstone: deploy your analytics assistant as an internal tool

Prerequisites

  • SQL proficiency
  • Basic Python
  • Access to a sample dataset or data warehouse

Outcomes

Your analytics team leaves with working tools that save hours per week, plus the skills to extend and customise them independently.

Interested in this masterclass?

Tell me about your team and I'll tailor the programme to your needs.

Book this masterclass