From 1,157 handwritten forms to one geological fact: a 25 m pile-driving limit

Transforming 1,157 handwritten forms into a geological fact: a 25m pile-driving limit. Discover how data extraction, aggregation, and resilient pipelines reveal hidden engineering knowledge.

Overview

A resumable extraction pipeline that turned 1,157 handwritten pile-driving field bulletins — scanned into 1,400+ messy PDFs — into a clean, georeferenced dataset, and then into engineering knowledge. The arc is the point: reports → data → knowledge.
Live, I’ll show the raw handwritten forms going in (smudged, non-standard, with voided sheets and unrelated pages mixed in), parallel subagents transcribing each to JSON, and the consolidated 1,157-record CSV driving an interactive report — histograms, box plots, an E×N heatmap, and a 3D scatter of 627 georeferenced piles. Then the payoff: those production-control reports became a technical and geological analysis instrument. Aggregated, the field data no single form could show revealed a consistent practical driving limit of ~25 m for the site (mean driven depth 24 m) — a hard geological fact that makes the next foundation estimate far more assertive. The dataset is a real completed job: 1,000+ pre-cast concrete piles driven over six months for a grain terminal in southern Brazil.

Tech stack

Claude Sonnet 4

Anthropic's mid-tier model: balances state-of-the-art coding (72.7% on SWE-bench) with cost-efficiency for high-volume, agentic workflows.

Claude Sonnet 4, released May 22, 2025, is Anthropic's versatile, mid-tier model (successor to Sonnet 3.7). It delivers superior coding performance, achieving a state-of-the-art 72.7% on SWE-bench. The model features hybrid reasoning: it offers both near-instant responses and an 'extended thinking' mode for deeper analysis. Optimized for high-volume applications, Sonnet 4 provides an optimal balance of capability and practicality, with pricing set at $3/$15 per million tokens (input/output). Access is readily available via the Anthropic API, Amazon Bedrock, and Google Cloud's Vertex AI.

https://www.anthropic.com/claude

View projects
Python

Python: The high-level, general-purpose language built for readability, powering everything from web backends to advanced machine learning models.

Python is the high-level, general-purpose language prioritizing clear, readable syntax (via significant indentation), ensuring rapid development for any team . Its ecosystem is massive: use it for robust web development with frameworks like Django and Flask, or leverage its power in data science with libraries such as Pandas and NumPy . The Python Package Index (PyPI) provides thousands of community-contributed modules, offering immediate solutions for tasks from network programming to GUI creation . The language is actively maintained by the Python Software Foundation (PSF), with the stable release currently at Python 3.14.0 (as of November 2025) .

https://python.org

View projects
Pandas

Pandas is the core Python library for high-performance data manipulation and analysis: it introduces the DataFrame and Series structures for fast, flexible handling of labeled data.

Pandas is your go-to, open-source Python library for data science, engineered for efficient data manipulation and analysis. It centers on two primary data structures: the two-dimensional DataFrame (like a spreadsheet or SQL table) and the one-dimensional Series. This toolkit handles everything from loading diverse file formats (CSV, Excel, JSON) to complex operations: data cleaning, transformation, statistical analysis, and time series processing. Developed by Wes McKinney in 2008, Pandas builds directly on NumPy and provides an intuitive API, making it the industry standard for turning raw data into actionable insights with minimal code.

https://pandas.pydata.org

View projects
Plotly

The leading open-source library for creating interactive, publication-quality data visualizations and web-based analytic applications.

Plotly is a technical computing company delivering powerful, interactive data visualization tools built on its core open-source library, Plotly.js. The platform supports multiple languages (Python, R, MATLAB) through wrappers like Plotly.py and Plotly.R. Its standout product is Dash: a framework for building full-stack analytic web applications entirely in Python, R, or Julia. This allows data science teams to deploy complex visualizations—from simple scatter plots to advanced 3D charts—directly to production environments, moving beyond static reports to deliver scalable data apps for Fortune 500 clients and millions of users.

https://plotly.com

View projects
HTML

HTML (HyperText Markup Language) is the foundational technology: it structures all web content using a system of elements and tags for browser display.

HyperText Markup Language (HTML) is the essential, foundational technology for the World Wide Web. It provides the core structure for all documents: defining elements like headings (`<h1>`), paragraphs (`<p>`), and critical hyperlinks (`<a>`). Browsers parse these tags to render a document into a complete multimedia web page. The current specification is maintained as a 'Living Standard' by the WHATWG, ensuring continuous, active development beyond the widely adopted HTML5 revision. HTML works in concert with CSS (for appearance) and JavaScript (for behavior) to deliver a complete, functional user experience across all platforms.

https://html.spec.whatwg.org/multipage/

View projects

From 1,157 handwritten forms to one geological fact: a 25 m pile-driving limit

Related talks