Portfolio

Reebooked

Developed an online platform and APP that allows users to buy and sell second-hand books,as well as donate books to NGOs in their local community contributed in the backend. • Built the project in the team of 4 people which was submitted to Google Solution Challenge,where I was the Team lead

Placement Predictor

Developed a placement predictor website which uses Machine learning(Random forest algorithm) • Built the project in the team of 3 people of GDSC ML Team, where I was the lead and My specific contributions included working on both the ML models and the frontend of the website.

Synthetic Poet

Developed Synthetic poets to recreate the work of Franz Kafka using Deep leaning and LSTM . • This model was trained on GPU for 2 days on the data of his famous book metamorphosis using tensorflow cuda library and generates synthetic poems in Kafka’s style by predicting the next word.

Leg Hill Climb Racing

Developed LegHillclimb Racing to learn the basics of Computer vision using Opencv,Mediapipe and python • Leg Hillclimb Racing can track the position of our ankles with relative position to our screen using Opencv and mediapipe which we can use to press brakes or slow down.

Portfolio Website

Web

Speech-tagging

Speech tagger utilizes probabilistic models using the Viterbi algorithm and Markov’s Chain to tag words in a text speech as nouns, pronouns, adverbs, etc.It was developed using Python • One of the features of this speech tagger is its ability to tag even ambiguous words using emission probabilities. The model takes into account the context of the word in the sentence.

Autocorrect

Python