Manasi Swaminathan

Experience

• 

  Swinburne University of Technology

  Dec 2018 - Jan 2019

  Research Trainee

  Conducted surveys for the development of smart cities and researched waste management and recycling methods as a part of design thinking Analyzed the learnings to design a product called iBin, an internet driven automated system for waste recycling and collection for IndiaInspired by the BigBelly bins and the current Industry 4.0 technologies like IoT and Cloud Computing

• 

  Amtex Systems Inc.

  May 2019 - Jun 2019

  Software Intern
• 

  CCCL India Ltd

  Oct 2019 - Dec 2019

  Software Intern
• 

  Indiana University Luddy School of Informatics, Computing, and Engineering

  Apr 2022 - now

  Researcher

  • Spearheaded a pioneering research project on developing a social robot for older adults, anchored in the concept of Ikigai, with Toyota Research Institute. This involved conceptualizing the project, engaging in rigorous design thinking, and executing comprehensive user studies.• Innovatively utilized cutting-edge Large Language Models (LLMs), including GPT-3.5, to implement advanced human-robot interaction mechanisms. This initiative enhanced the robot’s conversational abilities and user satisfaction, as evidenced by detailed user feedback analysis.• Intricately analyzed extensive target user interview data, discovering the personalized models ultimately developed to increase engagement and satisfaction within the same focus group measurably using quantitative and psychometric methods. Show less

• 

  FourKites, Inc.

  Aug 2023 - now

  Data Science Intern

  • Spearheaded FourKites’ Generative AI product research and design, bringing the product through development, Beta, and ultimately General Release with the quality and capability to draw over 380 recurring users across 40 accounts• Crafted a solution and quality assurance design that reduced negative sampling feedback by 74% (8.2% to 2.1%) before the first major version release. Enabled faster iteration cycles and increased feature quality.• Partnered with data science and engineering teams to build scalable Azure OpenAI architecture and design, increasing LLM development throughput by 30%, bringing the development back into line with the announcement, Beta and GA deadlines. Accelerated time-to-value.• Conceptualized and developed a proof of concept to showcase the viability of a platform model for their product direction. Using an LLM and proprietary APIs, successfully demonstrated to the executive leadership team the opportunity to lower development time by 90%+ and position a product with a $10M+ monetization opportunity to current accounts. Show less