Mohd. Zeb Alam

ChatBot

ChatBot for restaurant recommendations based on location and cuisine. A serverless app deployed using AWS lambda, with YELP API for restaurant data and fast storage API using ElasticSearch and Dynamodb. We used AWS Lex for language processing and AWS SNS for delivering notification.

Close Window

Smart door Application

Face recognition using opencv to detect if a person can enter a home or not. This was deployed using AWS Rekognition which picked data from AWS Kinesis Video stream. The face data was stored in a AWS Collections Record Db. SNS used for notification. New access was provided using a front end deployed for the main user

Close Window

Smart Photo Album

Photo album that performed image detection using AWS Rekognition. ElasticSearch is used for storing tagged data deployed over a VPC(Virtual Private Cloud). S3 event API used as a trigger. Voice Search performed using AWS Transcribe. (AWS Lambda, API Gateway, AWS lex, AWS Transcribe)

Close Window

Gofickle

A serverless web app for university shortlisting. Fast Storage using Elasticsearch and DynamoDb. Developed a feature to predict users ability to get accepted using AWS sagemaker on UCLA data using XGBoost Algorithm. Performed Data analytics on user click Stream data (using AWS Kinesis Stream, Amazon Firehose, AWS kinesis Data analytics, AWS crawler, AWS Athena, QuickSight).

Check out the video of the project

Close Window

Stock Market prediction

Implemented SVM classification approach for stock prediction on a data set of S&P 500 and New York stock exchange. Used RBF kernel, Linear Kernel, Polynomial Kernel,Sigmoid Kernel to compare efficiency with RBF kernel having highest efficiency of 73.9%.(Python)

Close Window

Emotion detection using Text

Developed a Python application that used various classification algorithms(Naive Bayes, Support vector machine, Logistic Regression) to detect 5 types (sadness, love, fear, anger, joy and surprise) of emotions using textual data scraped from Twitter. (Python)

Close Window