Role & Team

Position: Software Developer Intern
Team: AI
Location: Toronto, Canada (Remote)

Worked as a Software Developer Intern at Naturalytics, where I built AI-powered healthcare solutions. Designed a scalable patient onboarding system using Twilio, Dialogflow CX, and LLaMA 2, developed a speech-to-text pipeline with NVIDIA NeMo and Google Cloud Speech API, and created a Streamlit-based medical report assistant to simplify complex medical terminology for patients and non-technical users.

1 Internship

Naturalytics

Worked as a Software Developer Intern, building AI-driven healthcare solutions. Focused on patient onboarding automation, speech-to-text systems, and medical report simplification tools.

3 Projects

AI Healthcare Tools

Designed a scalable patient onboarding system with Twilio + Dialogflow CX, developed a speech-to-text pipeline with NVIDIA NeMo & Google Speech API, and built a Streamlit-based assistant to simplify medical reports for patients.

6 Skills

Technologies Applied

Twilio Programmable Voice, Dialogflow CX, LLaMA 2, NVIDIA NeMo, Python, Streamlit, Google Cloud Speech API, Git.

What I worked on

  • AI Onboarding System: Built a scalable patient intake solution using Twilio Programmable Voice, Dialogflow CX, and LLaMA 2 for automated call triage.
  • Speech-to-Text Pipeline: Engineered transcription workflow with NVIDIA NeMo, Python SpeechRecognition, and Google Cloud Speech API, achieving 85–90% accuracy across accents.
  • Medical Report Assistant: Developed a Streamlit app that translated complex medical reports into patient-friendly language using LLMs.
  • Collaboration: Worked closely with mentors to integrate AI models, optimize workflows, and document implementation steps for future scalability.
  • Impact: Improved healthcare accessibility by making patient onboarding and medical information more efficient and user-friendly.

Impact

  • Enhanced patient onboarding by automating voice-based triage with Twilio, Dialogflow CX, and LLaMA 2.
  • Improved transcription accuracy to 85–90% for multi-accent datasets using NVIDIA NeMo and Google Cloud Speech API.
  • Increased healthcare accessibility by simplifying medical reports into layman-friendly language with LLM-powered Streamlit tools.
  • Strengthened applied AI skills by integrating speech, NLP, and cloud APIs into production-ready solutions.
Ashwanth at Naturalytics Internship

Twilio Programmable Voice

Dialogflow CX

LLaMA 2

NVIDIA NeMo

Google Cloud Speech API

Streamlit

Python

01
Summary

Naturalytics — Highlights

Worked as a Software Developer Intern at Naturalytics (Toronto, Canada), contributing to AI-powered healthcare solutions. Built a patient onboarding system using Twilio Programmable Voice, Dialogflow CX, and LLaMA 2, engineered a speech-to-text pipeline with NVIDIA NeMo and Google Cloud Speech API, and developed a Streamlit-based assistant to simplify medical reports into layman-friendly explanations. Collaborated with cross-functional mentors to design, deploy, and document scalable solutions.

Role

Software Developer Intern — built AI-driven tools for patient onboarding, transcription, and accessibility.

Tech

Twilio, Dialogflow CX, LLaMA 2, NVIDIA NeMo, Google Cloud Speech API, Python, Streamlit, Git.

Contributions

Automated patient onboarding, improved speech-to-text accuracy, and created tools to make medical data more accessible.

Outcomes

Achieved 85–90% transcription accuracy, improved patient accessibility, and delivered scalable AI-driven healthcare solutions.