Machine Learning - Internship

You'll focus on training and optimizing computer vision and multimodal AI

models to analyze architectural and engineering drawings. This includes

dataset creation, model fine-tuning, and building pipelines that extract

structured data from visual technical documents.

Selected intern's day-to-day responsibilities include:

1\. Create and curate training datasets from architectural drawings, floor

plans, and technical specifications

2\. Annotate and label drawing elements (walls, doors, windows, dimensions,

symbols, text)

3\. Fine-tune vision models (GPT-4V, Claude Vision, or open-source

alternatives) for drawing analysis

4\. Train custom object detection models to identify specific drawing elements

5\. Experiment with OCR models specialized for technical drawings and

handwritten annotations

6\. Build evaluation frameworks to measure extraction accuracy across

different drawing types

7\. Optimize models for specific use cases (wardrobe specs, room layouts,

structural elements)

8\. Develop preprocessing pipelines for drawing normalization and enhancement

9\. Implement and test multimodal approaches combining vision and text

understanding

10\. Document model performance, training procedures, and best practices