Kausik Sivakumar

Kausik Sivakumar

Operations Manager Intern

Followers of Kausik Sivakumar1000 followers
location of Kausik SivakumarBoston, Massachusetts, United States

Connect with Kausik Sivakumar to Send Message

Connect

Connect with Kausik Sivakumar to Send Message

Connect
  • Timeline

  • About me

    Research Engineer @ NECLA | M.S.E. Robotics @ UPenn

  • Education

    • Amrita Vishwa Vidyapeetham, Coimbatore

      2016 - 2020
      Bachelor's degree Mechanical Engineering 9.37

      Recepient of 7th Rank amongst Mechanical Engineering students

    • University of Pennsylvania

      2021 - 2023
      Master's degree Robotics 3.91

      Activities and Societies: PAL Research Lab, DAIR research lab Recipient of the Outstanding Research Award for contributions toward Robot Learning/AI research

  • Experience

    • Amazon

      Feb 2020 - Jul 2020
      Operations Manager Intern

      Built Cost-Optimizing Delivery Forecasting Tool: Developed a statistical forecasting tool to estimate the required number of delivery vans, optimizing operational costs. This tool effectively reduced the company's delivery-related expenses by 5%.

    • Tata Consultancy Services

      Sept 2020 - Jul 2021
      Systems Engineer - Innovation labs

      Engineered Advanced Recommendation Systems: Developed sophisticated recommendation systems to model implicit feedback, including view duration and click counts. Leveraged Apache Spark to efficiently process and analyze large-scale datasets, optimizing the performance and scalability of the recommendation engine.

    • University of Pennsylvania

      Oct 2021 - Aug 2023

      Innovated Transition Occupancy Matching (TOM): Discovered and developed TOM, a novel policy-aware dynamics model learning method that focuses on transitions the policy can experience. This method drives the policy to achieve higher task rewards, leading to significant advancements presented at the L4DC 2023 conference and the NeurIPS 2022 Deep RL workshop. Advanced Goal-Conditioned Robot Manipulation: Led the initiative to develop a goal-conditioned robot manipulation policy by leveraging offline sub-optimal human video demonstrations, enhancing the robot's ability to learn and perform complex tasks effectively. Show less

      • Graduate Research Assistant - Reinforcement Learning

        May 2022 - Aug 2023
      • Graduate Teaching Assistant

        Jan 2022 - May 2022
      • Graduate Research Assistant - Physics Inspired Learning

        Oct 2021 - May 2022
    • Tutor Intelligence

      Aug 2023 - Oct 2024

      Pioneered Visual-Servoing Methodology: Developed and implemented a cutting-edge visual-servoing system that utilized image feedback to correct robot arm pick poses, achieving precision within 3mm and enhancing accuracy by over 40%.Optimized Robotic Trajectories: Applied advanced optimization techniques, including TOPPRA, to enhance the robot trajectory planning using Drake solvers. Innovated an online blending strategy that seamlessly integrated motion plans during robot operation, reducing pick-pick cycle time by 0.25 seconds.Enhanced Scene Understanding and Picking Precision: Fine-tuned the Segment Anything Model (SAM) to automate scene understanding, significantly improving the accuracy and efficiency of robotic picking tasks. Show less

      • Robotics Engineer

        Oct 2023 - Oct 2024
      • Robotics Intern

        Aug 2023 - Nov 2023
    • NEC Laboratories America, Inc.

      Oct 2024 - now
      Research Engineer
  • Licenses & Certifications

  • Honors & Awards

    • Awarded to Kausik Sivakumar
      Outstanding Research University of Pennsylvania May 2023 Award for Outstanding Research for contributions in Robot learning
    • Awarded to Kausik Sivakumar
      Ranked 7th in the Department of Mechanical Engineering Amrita Vishwa Vidyapeetham Aug 2020 Obtained the 7th rank (top 5 percentile) among my peers in the Mechanical Engineering undergraduate degree. Passed in First Class with a Distinction for the same.
    • Awarded to Kausik Sivakumar
      Finalist for Smart India Hackathon 2018 Government of India Feb 2018 This product was designed as a solution for Smart India Hackathon 2018.Worked with an amazing team in providing an solution to increase farmer profits by making a small scale automated supply chain system with root admin access. It was an app that would connect farmers and consumers directly, without losing the experience(negotiating price) that the customer would have while buying products face-to-face. This app wold effectively cut any middle men and automate the system to customers… Show more This product was designed as a solution for Smart India Hackathon 2018.Worked with an amazing team in providing an solution to increase farmer profits by making a small scale automated supply chain system with root admin access. It was an app that would connect farmers and consumers directly, without losing the experience(negotiating price) that the customer would have while buying products face-to-face. This app wold effectively cut any middle men and automate the system to customers. The app was done for different type of cottons for farmers in Maharashtra, as a prototype for the hackathon.We were one of the eight teams in the finals invited to present our app to ministry officials as part of our final pitch Show less
  • Volunteer Experience

    • Coding Instructor : Fife - Penn CS academy

      Issued by University of Pennsylvania on Oct 2021
      University of PennsylvaniaAssociated with Kausik Sivakumar