AI 1 day · In person
AI Course for Structural Engineers
For engineers and design firmsOne day to bring artificial intelligence into the design process: starting from AI chat and ending with verifiable working tools, built around the real workflows of a structural engineering practice.
Who it's for
- Freelance structural engineers who want to cut the time spent on repetitive tasks
- Design firms looking to modernize their processes and standardize internal know-how
- Designers using FEM software (Midas, SAP2000, ETABS…) who want to automate it
- No programming experience required
Program
Morning
The big picture · The Claude ecosystem
- LLMs applied to structural engineering
- Why Anthropic and Claude in a professional setting: reliability and privacy of project data
- An overview of the ecosystem: AI chat, Projects, Claude in Excel, Claude Code, Cowork
Working through project documents · Claude Cowork
- Extracting, aggregating and summarizing data spread across project documents: geological reports, material test certificates, existing drawings
- Critical review of calculation reports and results tables
- Creating "skills" to extend what an LLM can do on specific tasks
- Managing the LLM's context with .md files and subagents
- Hands-on exercise
Afternoon
Building a working tool · Claude in Excel
- Claude in Excel applied to the firm's verification spreadsheets
- Understanding and documenting inherited or uncommented spreadsheets
- From improvised sheet to real tool: structure, explicit formulas, code references
- Reusable templates carrying the firm's own styling, where the calculations are laid out in the format of a calculation report
- Consistency checks and testing against edge cases
- Guided exercise: building a complete verification sheet, from layout to validation
What you'll be able to do
- Extract, aggregate and cross-check data spread across project documents
- Create skills and manage the model's context for your practice's recurring tasks
- Build a reusable verification spreadsheet, laid out as a calculation report, without knowing how to program
- Critically validate the results produced by AI
Tailored deep dives
Further topics and deep dives can be requested for customized courses, run at your practice or company.
Building custom tools · Claude Code
- Claude Code: from zero to a working verification application, without writing code by hand
- From the terminal or inside an IDE such as Visual Studio Code, whichever suits you
- The development cycle: describing the problem, iterating, testing and validating
- Automating the firm's repetitive workflows
Integration with analysis software · Claude Code
- Creating custom tools for Midas Civil
- Generating and reading input/output files of FEM software
- Automatic extraction of results from analysis models
- Connecting to Rhino and Grasshopper for parametric design workflows
Standardization and validation · Projects and skills
- Encoding the firm's standards into reusable Projects and skills: report templates and checklists
- Strategies for validating results and managing the risk of error
- Confidentiality of project data and professional responsibility