Uneet Patel

B.Tech in Computer Science

Indian Institute Of Technology, Ropar



Programming Languages
  • C C++ Javascript Python Ruby Perl PHP
Libraries and Frameworks
  • ReactJS Mobx Redux Django Ruby on Rails Tensor Flow OpenCV
  • Postgres MySQL InfluxDB MongoDB

Tools And OS

  • MATLAB LateX PhotoShop After Effects

  • Linux MacOS Windows
Strong Hold Average Familiar

Exracurricular Activities and Responsibilities

  • Was an active member - Coding Club IIT ROPAR
  • Was a Head of Photography - Photography Club IIT Ropar
  • Active ENACTUS member and represented enactus IIT Ropar which stood 6th nationwide in Enactus National Competition held in Mumbai in 2015
  • Was an Active Member - video editing team, Cultural and Technical Fest of IIT Ropar
  • Represented IIT Ropar in Cricket at Inter IIT Sports Meet held in 2015 and 2016.

Work Experience

Founding Director & CTO - Voylr
Oct 2019-Present

Voylr is an online booking platform for in-destination activities and adventures. Our mission is to create a transparent marketplace in front of the customer by listing multiple authenticated field experts for the same service with the best possible prices. This helps us to create a customer-centric environment where no single customer should worry about negotiating to the best prices, safety, and reliability of the expert. Pay-on-site and it's one of a kind pricing strategy – customer-centric innovations that have made in-destination services booking more accessible and affordable for millions of Indians.

Summer Intern - Go-jek
May 2018 - July 2018

Developed an react native app that is used as a means for Live Data/Metric Visualization and Analysis Tool. The aggregated booking data/metrics with different topics was pushed into the INFLUXDB from KAFKA by an internal tool called DAGGER. The server side of the app was built in Ruby on Rails which provided RESTful API endpoints to access the aggregated time-series data/metrics from influxDB. The client side to visualize the aggregated data/metrics was built in REACT NATIVE independently from the server side. The back-end was containerized and deployed with CI/CD pipelines with the help of KUBERNETES. Final App provided the users a flexibility to create a Dashboard and visualize the desired data/booking-metrics of particular duration or days in the different form of a chart, bar-graphs, even simple numbers etc.

Relevant Course Project

Prediction of the Sentiment of a Movie Review Using ID3 Decision Tree Algorithm
Aug 2017 - Sept 2017

Stanford’s Large Movie Review Dataset containing 50k movie reviews was used. The file in the dataset contained around 90k words in which each word was used as attribute. The decision tree was learnt by ID3 algorithm, using the information gain heuristic to choose the next best attribute for splitting function to approximate the Occam’s Razor hypothesis. Various techniques like early stopping criteria to avoid overfitting, post-pruning strategy, and making a decision forest using feature bagging were used to increase the prediction accuracy on the test dataset.

Implemented and Analysed Advanced Spatial Filtering Techniques for Contrast Enhancements
Aug 2017 - Sept 2017

One of the most common Image enhancement technique Histogram Equalisation(HE) may result in over-enhancement and feature loss problems that lead to unnatural look and loss of details in the processed images. Therefore input image was processed through four levels of Intensity transformation namely Adaptive Histogram Equalisation(HE), Gap Adjustment (GA), Gray Value Recovery (GVR) and Dark region enhancement (DRE). The crux and details of above techniques were taken from the research article Contrast Enhancement Algorithm Based on Gap Adjustment for Histogram Equalization.

Functional And Pipeline Simulator for SimpleRISC Assembly Language
Sep 2016 - Nov 2016

SimpleRISC is a simple, generic, complete and concise RISC ISA. The assembly language of SimpleRISC has only 21 instructions in its ISA. Functional(with no pipeline technique) and Pipelined Simulator were developed in C++ producing the same output. Also compared the throughput differences between both of them. The pipelined simulator gave increased throughput due to better resource utilization. Both simulator contained five processor classes namely Instruction Fetch, Decode, Execute, Memory and Memory Write Back.

Universal Loudness Guard for Infrared Devices using Arduino and Infrared Sensors
Sept 2016 - Dec 2016

The device when paired with an audio system (supporting infrared signals) like television, speakers,etc. doesn’t allow the device’s volume to increase above an amplitude threshold (tunable by the user).If the sound amplitude increases above the set threshold, it is detected by the microphone and encoded infrared signals are sent to the device to decrease the volume until it is within specified limit.

Academics Details

Indian Institute Of Technology Ropar

B.Tech In Computer Science and Engineering

JEE(Joint Entrance Exam)Advanced

Conducted by Indian Institute Of Technology Bombay, AIR = 2149(GEN)

Divine Child High School
2013 - 2014

Centra Board Of Secondary Education Class XII (Percentage=90%)

L.P. Savani Vidhyabhavan
2010 - 2012

Gujarat Secondary Education Board Class X (Percentage=90%)

Gurukul Vidhyapeeth V.T. Choksi
2000 - 2010

Primary School