Nikhil Melgiri

Nikhil Melgiri  new_releases

About

Detail

Mechanical Engineering Student
Waterloo, Ontario, Canada
Contact Nikhil regarding: 
Flexible work
Starting at USD37.5/hour
id_card
Internships

Timeline

work
Job
school
Education

Résumé

Jobs verified_user 0% verified
  • Wysdom.Al
    Machine Learning Intern
    Wysdom.Al
    Sep 2020 - Current (5 years 10 months)
    • Manually labeled over 2,000 data points, decreasing the standard deviation of testing by a factor of 2. • Identified and corrected major codebase inefficiencies, decreasing model losses by 2x. • Created an NLP-based Neural Network capable of categorizing texts to corporate systems, eliminating the need for hundreds of hours of manual categorization, specifically using DistilBERT.
  • Staples
    AI/Robotics Intern
    Staples
    Jan 2020 - May 2020 (5 months)
    - Simplified and reduced core machine learning codebase over 10x enabling testing up to 50 neural networks per workday with automated statistic retrieval. - Manually labeled over 1,000 images via creating a GUI, decreasing the standard deviation of testing by a factor of 3. - Researched various neural networks to impliment in Staples for customer recommendations.
  • National Science and Engineering Research Council of Canada
    Artificial Intelligence Intern
    National Science and Engineering Research Council of Canada
    Jan 2019 - May 2019 (5 months)
    - Implemented an open-source machine learning algorithm into an autonomous vehicle vision lane marking system using data management, Python, and artificial neural networks (DNN) with GPU. - Invented new LiDAR point cloud filters to reduce the data processing time by over 90%, enabling real-time analysis with LiDAR and SLAM (simulation localization and mapping). - Participated in researching self-driving vehicles.
  • Intelline Inc
    Engineering Intern
    Intelline Inc
    Aug 2018 - Sep 2018 (2 months)
    • Decreased total base volume by over 30%, enabling high capacity storage of cryocooling systems. • Decreased total base weight by over 40%, fulfilling the company's goal of making cryocoolers marketable for hobbyists and low-budget programs • Increased tolerances and precision, shaving 45% off of milling time (including time reduction from lower volume). • Conducted major repairs on the CNC machine, increasing personal knowledge in CNC systems.
  • C
    Design Sprint Intern
    Conrad School of Entrepreneurship and Business
    Jul 2018 - Aug 2018 (2 months)
    • Conducted a series of design sprints to help entrepreneurs overcome particular obstacles • Increased user retention by approximately 20% in SheLeads by designing new games and collaborating with students and instructors • Made UX for instructors vastly simplified for SheLeads - Created a website framework for Notonian
  • Sienci Labs
    Engineering Intern
    Sienci Labs
    Jul 2018 - Aug 2018 (2 months)
    • Enhanced the CNC sensory systems by over 10x by optimizing a platform for a piezoelectric sensor. • Vastly increased hardware durability of core computer components from months to years by redesigning the case • Increased packaging rate of components by over 3x from designing a custom 3d printed specialized spool • Designed and Lathed couplers and pulleys to assist in prototyping a new CNC machine version • Improved marketability by milling out a model rocket
Education verified_user 0% verified
  • University of Waterloo
    Mechanical Engineering
    University of Waterloo
    Sep 2017 - Current (8 years 10 months)