ReSolveAI - AI Complaint Classification Engine 🧭🤖

Automated Complaint Understanding for Financial Services
ReSolveAI transforms unstructured customer complaints into actionable, categorized insights using NLP, topic modeling, and supervised machine learning.
This project blends practical engineering with research-grade experimentation.

📄 Full Report & Slides included (see Resources section).

🔍 Overview

ReSolveAI is an end-to-end AI engine that:

• Cleans and preprocesses raw complaint text
• Extracts meaningful patterns using TF-IDF
• Discovers hidden topics with Non-Negative Matrix Factorization (NMF)
• Uses these topics to train classification models
• Produces automated complaint category predictions with explainability

Why ReSolveAI matters:

• ⚡ Faster resolution
• 🎯 Higher routing accuracy
• 💸 Reduced operational overhead
• 😊 Improved customer satisfaction

This project analyzed 78,313+ complaint records across 22 metadata fields, using label-aligned topic clusters from NMF.
:contentReference[oaicite:2]{index=2}

---

📁 Repository Structure

├─ README.md

├─ src/

│ ├─ notebook.ipynb

├─ dataset/

│ ├─ complaints-2021-05-14_08_16.json

├─ docs/

│ ├─ results/

│ │ ├─ decision-tree.png

│ │ ├─ gaussian-naive-bayes.png

│ │ ├─ logistic-regression.png

│ │ ├─ random-forest-classifier.png

│ ├─ architecture.png

│ ├─ workflow.png

│ ├─ wordcloud.png

│ └─ n-gram.png

├─ report.pdf

└─ presentation.pdf

---

✨ Key Features

• 🔤 NLP Pipeline: tokenization, lemmatization, POS tagging, stopword filtering
• 🔍 Topic Modeling with NMF – reveals semantic clusters
• 🧮 Multiple ML Classifiers
  • Logistic Regression (≈95% accuracy)
  • Decision Tree
  • Random Forest
  • Gaussian Naive Bayes
    (LR was the top performer.)
• 🎨 Rich EDA: histograms, distributions, n-grams, word clouds
• ⚙️ Complete Inference Pipeline: ready for deployment
• 🧪 Interpretability: TF-IDF feature importance, topic keywords
• 📚 Full research documentation in PDF

---

🏗 System Architecture

Architecture: Ingestion → Processing → Topic Modeling → ML Classifier → Prediction

---

🔄 Workflow

Model Workflow: Data → Clean → TF-IDF → NMF → Classifier → Results

---

📊 Exploratory Data Analysis

Word Cloud of Most Frequent Complaint Tokens

From the dataset EDA:

• Strong presence of tokens like payment, credit, account, dispute, reporting
• Complaints map naturally into financial service categories
• Topic-token alignment validated using NMF factors
  :contentReference[oaicite:3]{index=3}

---

🚀 Getting Started

1️⃣ Clone the repository

git clone https://github.com/kalpthakkar/JobPilot-AI.git
cd jobpilot_ai

2️⃣ Create Python Environment

python -m venv .venv_resolveai

# macOS/Linux
source .venv_resolveai/bin/activate


# Windows
.venv_resolveai\Scripts\activate
# Powershell
.venv_resolveai\Scripts\Activate.ps1

3️⃣ Run Notebook

jupyter lab src/notebook.ipynb

---

🧠 Example: End-to-End Model Pipeline

from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.decomposition import NMF
from sklearn.linear_model import LogisticRegression
from sklearn.model_selection import train_test_split
from sklearn.metrics import classification_report

tfidf = TfidfVectorizer(max_df=0.95, min_df=2, ngram_range=(1,2))
X = tfidf.fit_transform(df['clean_complaint'])

nmf = NMF(n_components=5, random_state=42)
W = nmf.fit_transform(X)

y = W.argmax(axis=1)

X_train, X_test, y_train, y_test = train_test_split(
    X, y, test_size=0.2, random_state=42
)

clf = LogisticRegression(max_iter=1000)
clf.fit(X_train, y_train)

print(classification_report(y_test, clf.predict(X_test)))

---

📈 Results Summary

Overall Model Performance

Model	Accuracy	Notes
Logistic Regression	⭐ ≈95%	Best performer
Decision Tree	~77%	Interpretable
Random Forest	~72–74%	Stable but overfit-prone
Gaussian NB	~36%	Weak baseline

• ✔ LR chosen as final classifier
• ✔ Well-separated NMF topics improved class clarity
• ✔ Topic-word alignment validated

---

🧩 Future Enhancements

• Fine-tuned transformer models (GPT, BERT, DistilBERT)
• Real-time ingestion with Kafka
• FastAPI inference service
• Active learning loop from agent feedback
• Multi-modal complaint classification (voice → transcripts → NLP)

📚 Resources

📄 Full Research Report (PDF)

📑 Project Presentation (Slides)

💬 Frequently Asked Questions (FAQ)

Q: What dataset did you use?
A dataset of 78,313 customer complaints, with 22 columns, provided in JSON format.

Q: Why NMF for topic modeling?
NMF provides sparse, interpretable topics that align well with complaint categories.

Q: Why did Logistic Regression perform best?
High-dimensional TF-IDF vectors naturally favor linear decision boundaries.

Q: Can this model run in production?
Yes, TF-IDF + LR is fast, light, and easily containerizable.

Q: How is bias or PII handled?
Remove PII (names, IDs, emails) and evaluate demographic fairness before launch.

🧾 Citation / Acknowledgements

This Machine Learning work is developed by Kalp Thakkar. References and full experiments in report.pdf and slides

Bird, Klein & Loper - Natural Language Processing with Python (NLTK).
Pedregosa et al. - Scikit-learn.
Lee & Seung - NMF (Nature, 1999).

See full bibliography in this report

🤝 Contribution

• Fork repository

  https://github.com/kalpthakkar/ReSolveAI-AI-Complaint-Classification-Engine/fork

• Create a branch:

  git checkout -b feature-xyz

• Commit your changes

  # Stage changes
  git add .
  
  # Commit
  git commit -m "Your message"
  
  # Push the new branch
  git push -u origin feature-xyz

• Submit a Pull Request

  gh pr create --fill

---

❤️ Final Note

ReSolveAI showcases how classical NLP + topic modeling + supervised ML can deliver real, measurable impact in customer complaint handling pipelines. This repository demonstrates research quality, engineering quality, and practical industry relevance - all in one project.

---

📞 Contact

For any inquiries or support, please contact:

• Kalp Thakkar - kalpthakkar2001@gmail.com
• GitHub: kalpthakkar
• LinkedIn: kalpthakkar