AI · AIC-16

Claude AI for Data Analyst

Claude AI for Data Analyst is a course in taking data from raw files to insight with Claude AI, covering data-focused prompts, KPI design with the GQM framework, data cleansing, Claude Code with Python and Pandas, and interactive dashboards. It suits data and business analysts, managers who make data-driven decisions and SME owners without a large data team.

Updated 21 Aug 2026

From 8,010 THB / person 8,900 −10% excl. VAT 7% · group rates available

Corporate training quote Join the waiting list

PDFDownload the course outline

Duration12 hours · 2 days
FormatOnsite / live online
Next roundOn request
CertificateIncluded

Course overview

Every organization now holds a vast amount of data, so the real problem is not a lack of data but turning it into decisions fast enough. Many analytics teams still spend most of their time on repetitive work such as cleaning data, formatting tables, building summary slides and answering the same executive questions every month, leaving less and less time for deep analysis. This course upgrades data-analysis skills with Claude AI, focusing on real work from raw data all the way to insight and decision. Learners cover choosing the right model for the task, designing prompts specifically for data work, setting goals and KPIs with the GQM Framework,

data cleansing and ETL, reproducible analysis with Claude Code and Python/Pandas, designing dashboards that answer business questions, building interactive dashboards and live artifacts, writing executive insight reports, and finally setting up automation with Claude Cowork, Claude Skills and MCP Connectors to systematically cut the time spent on routine reports. The highlight of this 2-day course is using Claude Code together with Python/Pandas so results are reproducible, creating Claude Skills so monthly reports become standardized automated work, and putting data governance and security in place in line with PDPA. All content is taught through workshops using realistic sample datasets and finishes with an end-to-end capstone project

What you’ll gain

Understand the capabilities and limits of Claude AI for data analytics and choose the right model for the task
Design prompts specifically for data work using the Ask, Analyze, Validate, Visualize and Recommend patterns
Understand data thinking and design KPIs with the GQM Framework (Goal, Question, Metric) tied to real organizational goals
Check data quality, plan data cleansing and run ETL with Claude AI systematically
Use Claude Code with Python and Pandas to build reproducible analysis that can be traced and repeated
Design dashboards that answer business questions and build interactive dashboards, HTML reports and live artifacts
Write executive insight reports and a morning insight brief using the what happened, why, what next structure
Apply Claude Cowork, Claude Skills and MCP Connectors to automate routine reporting
Set data governance, security and PDPA practices for using AI with organizational data
Assess the accuracy of AI-generated results and verify the numbers before using them to decide

Who this course is for

Data analysts and business analysts who want to cut reporting time and deepen their analysis
Managers, team leads and executives who read and decide from data regularly
Business owners and SME entrepreneurs who want data-driven decisions without a large data team
Staff in planning, finance, marketing, operations and HR who prepare monthly summary reports
Those interested in becoming a data analyst who want to use AI to accelerate learning

Prerequisites

Basic computer and internet skills
Basic Microsoft Excel or spreadsheet skills such as opening files, managing data and reading data tables
Some experience working with data such as reporting, summarizing or basic analysis gives the most benefit
Understanding of your own team's workflow or business goals to use as a case study during training
No programming or data science background required

Curriculum

Course Details

A 2-day course, 6 hours per day (12 hours in total), delivered as a hands-on workshop with a capstone project. Intermediate level with no programming background required. Content is taught with realistic sample datasets (sales, customer, operations and finance). Learners take home a prompt library, a Claude skill, an executive report template, a dashboard wireframe and a data governance and PDPA checklist.

Day 1 From Raw Data to Trustworthy Analysis

Section 1: Claude AI as Your Data Analyst Assistant

Overview of generative AI for data analytics, choosing the right model (deep reasoning, high-volume analysis, report writing, file generation) and understanding the context window for large data
Uploading and working with many file formats (CSV, Excel, PDF, JSON, images) and using Projects and Project Knowledge to reuse organizational context
Limits of AI on numeric work (hallucination, miscalculation, misreading tables) and a verification mindset before using numbers
Workshop: upload a sample sales dataset and have Claude summarize it while you verify the numbers

Section 2: Prompt Engineering for Data Analysts

A good prompt structure for data work (Role, Context, Data, Task, Format, Constraint) and how it differs from general prompts
Five key prompt patterns: Ask, Analyze, Validate, Visualize and Recommend, plus chain-of-thought and few-shot examples
Building a personal prompt library for monthly recurring work
Workshop: design a prompt set for a report you actually produce, then test and refine it

Section 3: Data Thinking with the GQM Framework

Data analytics thinking that starts from the question not the data, and the levels of analysis: descriptive, diagnostic, predictive and prescriptive
The GQM Framework turning business goals into measurable KPIs, the difference between KPI, metric and dimension, and leading vs lagging indicators
Choosing metrics that do not fool you: vanity metrics, survivorship bias and Simpson's paradox
Workshop: build a GQM Canvas for your own team and have Claude check the KPI logic

Section 4: Data Cleansing and Data Preparation with Claude

Having Claude do data profiling and detect issues: missing values, duplicates, outliers, inconsistent formats and wrong data types
Thai-specific data issues (Buddhist vs Gregorian dates, Thai numerals, extra spaces, inconsistent name spelling) and a data cleaning plan that records what was changed
Lookup, merge, join, grouping and calculated columns via natural language, plus managing a data dictionary
Workshop: clean a real customer dataset from inspection to an analysis-ready file

Section 5: Reproducible Analysis with Claude Code and Python/Pandas

The problem with in-chat analysis (inconsistent results, hard to trace) and the concept of reproducible analysis
Having Claude write Python and Pandas scripts for cleansing and aggregation, and building a small ETL pipeline that reruns every month
Writing data validation checks to catch anomalies automatically and exporting results as Excel with real formulas
Workshop: build a monthly sales analysis script that takes a new file and produces a report instantly

Section 6: Day 1 Wrap-up and Review Workshop

Review the workflow: frame the question, prepare data, analyze and verify, plus common Day 1 mistakes and how to avoid them
Workshop: use the cleaned dataset to answer three business questions from your own GQM Canvas and prepare for Day 2

Day 2 From Insight to Dashboard, Report and Automation

Section 7: Dashboard Design and Insight Generation

Dashboard design principles for executives: a dashboard must answer questions not just show charts, and the strategic, operational and analytical types
Choosing charts that fit the question, visual hierarchy, meaningful use of color and presentation traps that mislead
Having Claude find insights, anomalies, trends and correlations, and distinguishing correlation from causation
Workshop: design an executive dashboard wireframe from the Day 1 dataset

Section 8: Claude Artifact and Interactive Dashboard Prototype

Building an interactive dashboard prototype with Artifact without coding, adding filters, drill down, sorting and search
Creating HTML reports and live artifacts that pull fresh data and refresh, plus tuning the design system to the organization brand
Where Artifact fits and when to move to Power BI or a full BI tool
Workshop: build an interactive dashboard from your wireframe and test it in use

Section 9: Executive Insight Report and Research Workflow

A report structure executives can act on within three minutes using the what happened, why, what next narrative and the Pyramid Principle
Using Claude Research and web search for market and competitor data, with source citation and credibility checks
Workshop: write a one-page executive insight report from your own analysis

Section 10: Automation, Claude Cowork and MCP Connectors

Assessing routine work for what to automate and what always needs a human, and using Cowork to handle many files on the user machine
An overview of MCP (Model Context Protocol) and connecting common connectors (Google Drive, Google Sheets, Gmail, MySQL, PostgreSQL)
Creating a morning insight brief and scheduling daily, weekly and monthly reports
Workshop: set up one automation to summarize data from a folder of report files

Section 11: Claude Skills for Recurring Organizational Reports

The structure of a Skill (SKILL.md, a good description, supporting files) and writing a description so the Skill triggers at the right time
Embedding report templates, calculation standards and business rules into a Skill so the monthly sales report is identical every time
Sharing Skills across the team and an overview of the Claude Agent SDK for full automation (demo)
Workshop: build one Skill for a report you must repeat every month

Section 12: Data Governance, Security and PDPA

Data classification, personal data under PDPA and what must never be entered into AI systems
Risk-reduction techniques: data masking, anonymization, pseudonymization and removing unnecessary columns
Setting connector and MCP permissions on least privilege, and establishing an AI usage policy, audit trail and human in the loop
Workshop: inspect a sample dataset, identify sensitive data and draft your team's practice guidelines

Section 13: Capstone Project and Real-world Adoption

Task: build an end-to-end analysis from a given business dataset or your own team data, setting goals and KPIs with the GQM Framework
Clean and analyze the data to find three key insights, build an interactive dashboard with Claude Artifact and write an executive insight report with recommendations
Present the work and get feedback, plus a first-30-days adoption plan and a learning path (Power BI, SQL, Python for Data Analysis, enterprise AI agents)

Schedule & training options

For individuals — public rounds

No public rounds are open right now. Join the waiting list and we will contact you first when the next round opens, or ask us on LINE. Or call 02-570-8449 or 088-807-9770

Join the waiting list Ask about rounds

For organisations — in-house / private

Tailor the content to your team’s tools and projects
Your dates, at your office or live online
Quotation with tax ID for procurement

Corporate training quote

Instructors

Samit KoyomFounder & CEO, IT Genius Engineering

Frequently asked questions

What is the GQM framework and how do you use it to design KPIs?

GQM stands for Goal, Question, Metric. Start with a goal linked to the organisation's objectives, such as increasing repeat sales, then write the questions you need answered to know whether you are on track, such as what share of customers buy again. Finally choose metrics that answer each question from the data you have. This avoids KPIs that are easy to measure but do not help decisions.

How do you check that numbers produced by AI analysis are correct?

AI can miscalculate, misread tables or invent figures, so ask it to show its method and formulas, reconcile totals against the source with Excel or a PivotTable, spot-check a few rows by hand, and have it work through code such as Python and Pandas so the analysis can be traced and rerun. Any figure used for a decision should be confirmed by a person.

What are the steps to clean data with Claude?

Upload the file and ask Claude to profile the data quality first: blanks, duplicates, inconsistent date formats and outliers. Have it propose a plan for each issue, review and approve the rules, then let it clean the data and report what changed and in how many rows. Finally compare row counts and totals with the original, and save the steps as a prompt or code to reuse next month.

Who is the Claude AI for Data Analyst course for, and what should I know beforehand?

It suits data and business analysts who want to cut reporting time, managers and executives who decide from data, SME owners without a data team, and planning, finance or marketing staff who prepare monthly reports. You should have basic Excel or spreadsheet skills, and some experience producing reports or summaries will help you get the most from it. No programming or data science knowledge is needed.

What will I be able to do after the Claude AI for Data Analyst course?

You will be able to write data prompts using the Ask, Analyse, Validate, Visualise and Recommend patterns, design KPIs with GQM, check and clean data quality, use Claude Code with Python and Pandas for reproducible analysis, build dashboards and HTML reports, write executive insight reports, and automate routine reporting. Participants who complete the training receive a Certificate of Completion with its own number, which can be verified online.

Can the Claude AI for Data Analyst course be run in-house for our team?

Yes. It can be run in-house for your team, with examples adapted to your own work, such as using the monthly reports and KPIs your team owns as exercises, along with data and PDPA guidelines for your organisation. You can request a quotation from this page, or contact us on LINE @itgenius or by phone on 02-570-8449.

Articles are published in Thai.