---
title: "Discussion on \"Building CivicAI  Part 1: \nWhen No Dataset Exists, You Build One\" | Hashnode"
description: "Discussion on \"Building CivicAI  Part 1: \nWhen No Dataset Exists, You Build One\". How I built a 1,100-row dataset from scratch for an AI municipal complaint system, formulas, labels, and preprocessing explained."
url: "https://hashnode.com/posts/building-civicai-part-1-when-no-dataset-exists-you-build-one/6a04c79ea0c1540277687d08"
---

[Feed](https://hashnode.com/feed) /

Discussion

# [Building CivicAI Part 1: When No Dataset Exists, You Build One](https://ai-ml-datascience-from-scratch.hashnode.dev/building-civicai-part-1-when-no-dataset-exists-you-build-one?utm_source=hashnode&utm_medium=feed)

How I built a 1,100-row dataset from scratch for an AI municipal complaint system, formulas, labels, and preprocessing explained.

May 13 · 11 min read · ai-ml-datascience-from-scratch.hashnode.dev  
Read the article

-   [#ml](https://hashnode.com/tag/ml)
-   [#ai](https://hashnode.com/tag/ai)
-   [#dataset](https://hashnode.com/tag/dataset)
-   [#tf-idf](https://hashnode.com/tag/tf-idf)
-   [#data-preprocessing](https://hashnode.com/tag/data-preprocessing)
-   [#nlp](https://hashnode.com/tag/nlp)
-   [#python](https://hashnode.com/tag/python)
-   [#data-science](https://hashnode.com/tag/data-science)
-   [#beginners-learningtocode-100daysofcode](https://hashnode.com/tag/beginners-learningtocode-100daysofcode)

## Written by

[wajahat taj](https://hashnode.com/@Wajahattaj)

@Wajahattaj · islamabad

  
[View profile](https://hashnode.com/@Wajahattaj)

Software Engineering student at FAST-NUCES University.

Software Engineering student at FAST-NUCES building full-stack AI-powered systems end-to-end. I work across the full stack: Frontend: React.js &amp; Next.js Backend: FastAPI &amp; Express.js Mobile: React Native ML &amp; NLP: scikit-learn, TF-IDF, Logistic Regression, Random Forest , XG-Boost , ANN. Deployment: Docker, Render, Vercel I don't just study ML concepts — I integrate them into real products and ship them. Currently going deeper into backend architecture, scalable API design, MLOps, and production deployment pipelines. Because writing code is just the beginning — shipping systems that work in the real world is the goal.

## Responses

Markdown supported · ⌘↵ to post

No responses yet.
