Akshat Tripathi
AI & ML engineer building production-ready impact models
New to Contra
Akshat is ready for their next project!
I have completed the E commerce sales forecasting project about which i posted yesterday.
Developed it using Python and some of its libraries like Scikit-Learn, Pandas, etc. Deployed it on the streamlit cloud through which anybody can use it.
In this you have to fill in some of the details about the product like product category, price, discount percentage, etc. After getting this details the model will then predict the sales according to the details.
The Accuracy of the model is approx 96% which makes it a very trust worthy model for your E commerce business.
1
44
💬 Customer Sentiment Analyzer
NLP-Based Review Classification | Python | Streamlit
I developed a Customer Sentiment Analyzer that classifies reviews as Positive or Negative using Natural Language Processing and a Logistic Regression model built with scikit-learn.
The system extracts data from CSV files using Pandas, preprocesses text, converts it into numerical features through vectorization, and delivers accurate sentiment predictions through an interactive Streamlit web app.
Tech Stack: Python, Pandas, NLP, scikit-learn, Logistic Regression, Streamlit
Designed to help businesses quickly analyze customer feedback and make data-driven decisions.
1
30
🥇 Gold Price Prediction – Machine Learning Web Application (Python | Streamlit)
I developed a Gold Price Prediction system that forecasts future gold prices using historical market data and machine learning algorithms. This project helps investors and analysts make data-driven decisions by identifying trends and patterns in gold price movements.
🔹 Project Features
📈 Predicts gold prices based on historical financial data
🤖 Machine learning model built using scikit-learn
📊 Data preprocessing, feature engineering, and model evaluation
📉 Performance metrics such as R² Score and Mean Squared Error
🌐 Interactive web application built using Streamlit
🔹 Technology Stack
Python – Core implementation
scikit-learn – Regression modeling
Matplotlib / Visualization tools – Trend analysis
Streamlit – Web application deployment
2
3
77
📰 Fake News Detector – NLP & ML Web Application (Python | Streamlit)
I developed a Fake News Detection system that identifies whether a news article is real or fake using Natural Language Processing (NLP) and machine learning classification techniques. The project analyzes textual patterns, word usage, and linguistic features to make accurate predictions.
🔹 Project Features
🧠 NLP-based text preprocessing and feature extraction
🤖 Classification model built using scikit-learn
📰 Detects and classifies news as Fake or Real
🌐 Interactive web application developed with Streamlit
🔹 Technology Stack
Python – Core implementation
scikit-learn – Machine learning & classification algorithms
NLP techniques – Text cleaning, tokenization, vectorization
Streamlit – Web app development and deployment
1
44
⚡ Electricity Bill Predictor – ML-Based Web App (Python | Streamlit)
I developed an Electricity Bill Predictor that estimates monthly electricity costs based on the usage patterns of household electrical appliances. This project helps users understand and manage their power consumption more efficiently by providing data-driven bill predictions.
🔹 Project Highlights
📊 Predicts electricity bills based on appliance usage inputs
🤖 Machine learning model built using scikit-learn
🧮 Processes usage data such as duration, frequency, and appliance type
🌐 Interactive web application developed using Streamlit
⚡ Fast, simple, and user-friendly interface
🔹 Technology Stack
Python – Core programming and logic
scikit-learn – Machine learning model for prediction
Streamlit – Web app UI and deployment
1
38
Freelance Project Description: Aluminium Wire Rod Casting Property Prediction
As a freelance machine learning developer, I designed and deployed a custom predictive system to estimate aluminium wire rod casting properties using existing production parameters. The objective of this project was to help the client improve product quality, reduce trial-and-error in manufacturing, and optimize casting decisions through data-driven insights.
The solution was implemented in Python, utilizing Pandas for data cleaning, preprocessing, and feature engineering, and Scikit-learn for building and training the machine learning model. The final model achieved a prediction accuracy of 97%, ensuring high reliability and consistency in real-world usage.
0
33
I developed Agribot, an AI-powered chatbot designed specifically to assist farmers with day-to-day farming and agriculture-related queries. The goal of this project is to make technology accessible to farmers in a simple, fast, and language-friendly way.
🔹 Key Features
🎙️ Voice-to-Text Support – Farmers can speak their questions instead of typing
⌨️ Text-Based Chat – Easy-to-use typing interface for queries
🌾 Agriculture-Focused Assistance – Crop issues, farming practices, and related guidance
🌐 Multi-Language Ready – Currently built in Hindi, with the capability to expand into multiple regional and global languages
⚡ Clean & Interactive UI – Built using Streamlit for a smooth and user-friendly experience
🔹 Tech Stack
Python – Core logic and chatbot intelligence
Streamlit – Frontend UI for fast and interactive deployment
0
35