Naman Singh

Computer Science and Applied Mathematics Undergraduate @IIIT Delhi

New Delhi, Delhi, India

About

Experience

  • Founder at CrescoCare
    Jan 2024 - Present · 2 yrs 6 mos

    Cresco Care was born out of a belief in collective growth—our name itself means "to grow together." Through grassroots initiatives, awareness campaigns, and targeted support programs, we aim to uplift underserved communities, provide access to essential resources, and empower individuals with the tools they need to lead better lives. From organizing health camps and educational drives to supporting local livelihood efforts, our work is rooted in compassion, sustainability, and real-world impact. We believe that meaningful change begins with empathy and action—and we’re here to be that catalyst.

  • Undergraduate Researcher at TAV Lab (AI for Medicine and Public Health)
    Jan 2025 - Jan 2026 · 1 yr 1 mo

    Undergraduate Researcher at TAV Lab (AI for Medicine and Public Health), IIIT Delhi, working on a Drowsiness Detection and Driver Safety System. Focused on real-time facial monitoring, drowsiness detection, and hardware integration to enhance road safety.

  • AI Model Trainer at Outlier
    Nov 2024 - Jan 2026 · 1 yr 3 mos

    Working as a Freelancer at Outlier, specializing in training and fine-tuning AI models to enhance their performance and accuracy. Passionate about leveraging cutting-edge machine learning techniques to drive impactful AI solutions.

  • AI Intern at SPIRO
    May 2025 - Jul 2025 · 3 mos

    Shipped LLM-powered customer support and speech systems. Designed and deployed WhatsApp chatbots using LLMs and RAG pipelines. Built a meeting summarization system using transformer models for faster team handoffs. Developed a real-time graph visualizer for live metric monitoring. Engineered a multilingual voice cloning TTS pipeline for Indian languages using open-source speech models

  • Undergraduate Researcher at Complex Systems Laboratory
    Jan 2025 - May 2025 · 5 mos

    We developed a model for dish detection by scraping and annotating images, and trained it using the YOLO (You Only Look Once) architecture.