Ishan Chugh

Experience

• 

  GAMER.XYZ

  Sept 2022 - May 2023

  Community Manager

  Managing community platforms like discord for https://arena.gl/ and running tons of events throughout the year for the audiences. Development of Discord Bots and other softwares.

• 

  AlmaBetter

  Jan 2023 - Mar 2023

  Discord Developer

  Built a Discord Community and a Discord Bot on Python for AlmaBetter, which included various features like community onboarding, quizzes, leaderboards, etc.

• 

  Smart India Hackathon

  Aug 2023 - Dec 2023

  Winner

  Winner of Smart India Hackathon 2023, excelling in the theme of Social Media Analytics and Tracking. Collaborating with a talented team, we devised an innovative solution for advanced social media monitoring known as SocioTrackr. Proudly emerged as the national winners in the competition.

• 

  AIMS-DTU

  Sept 2023 - now

  Co Head

  As the Cohead of AIMS-DTU, I'm proud to be part of a student society uniting DTU's AI and machine learning enthusiasts. Our mission is to foster AI and machine learning research at DTU, with seniors guiding juniors on their ML journey. In my role as the Cohead of the AI Team, I lead multiple projects and research initiatives, shaping the future of AI at AIMS-DTU.

• 

  Goldsmiths, University of London

  Oct 2023 - Jan 2024

  Research And Development Intern

  During my research internship at Goldsmiths, University of London, I delved into the fascinating world of sentiment analysis and explored the potential of novel NLP architectures. My focus was on enhancing the understanding of user-generated content, moving beyond simple sentiment detection to a deeper level of comprehension – Natural Language Understanding (NLU). This shift towards NLU marked a significant step in my research journey, opening doors to explore the capabilities of Large Language Models (LLMs).One of the key areas I investigated was the development of NLP architectures using natural language inference and fusion models. By combining these techniques, we aimed to achieve a more nuanced understanding of the underlying emotions and intentions within text data. I was particularly intrigued by the ability of LLMs to capture contextual information and identify subtle linguistic cues that traditional NLP models often miss. I experimented with various LLM architectures and fine-tuned them on the MBTI dataset, achieving an impressive accuracy of 88% in predicting personality types based on text samples.This internship was a stepping-stone in my exploration of NLU and LLMs, igniting a passion for developing intelligent systems that can truly understand and respond to human language. Show less

• 

  Deepsolv

  Mar 2024 - now

  AI Engineer Intern

  As an AI Engineer, I lead initiatives to advance AI-driven customer support agents, focusing on the development and optimization of innovative AI technologies. My role centers on:• Large Language Model (LLM) Optimization: Enhancing LLM agents for improved customer service response accuracy and conversation quality.• Speech Recognition Improvement: Upgrading speech recognition systems to ensure accuracy and ease of communication between customers and AI agents.• Data Organization with RAG: Implementing Retrieval-Augmented Generation (RAG) to streamline data access and retrieval, significantly improving response efficiency.• Latency Reduction: Actively minimizing system latency with LLM-based strategies, facilitating faster and more reliable user interactions.In this role, I am committed to leveraging AI for creating meaningful, user-centric solutions that not only solve technical challenges but also enhance the overall customer experience in support systems. Show less