
Case study
AI-powered medical scribe SaaS
An AI medical scribe that turns live consultations into structured clinical notes, letters and referrals in seconds.
Overview
- Client
- Australian healthtech startup
- Location
- Australia
- Industry
- Healthcare
An Australian healthtech startup wanted to free clinicians from late-night charting with an AI scribe that writes the notes while they consult.
They engaged us to build a secure, cross-platform product for a wide range of healthcare professionals, end to end.

The challenge
The AI had to handle complex medical terminology accurately and be secure enough for clinical use.
It needed to transcribe live consultations and telehealth sessions and turn them into structured clinical notes, without getting in the practitioner's way.
The solution
What we built
We built and tuned an AI scribe platform designed specifically for clinical settings.
Live transcription
Secure audio processing that transcribes consultations in real time.
Notes in clinical formats
Generative AI turns transcripts into SOAP notes, letters and referrals, using built-in or fully custom templates.
Insights and telehealth
AI insights for practitioners, with support for remote telehealth consultations.
Built for compliance
Encryption and controls aligned with HIPAA, the Australian Privacy Principles and SOC 2 Type 1.
The outcome
What changed afterwards
- Less time on paperworkClinicians spend far less time on manual charting and more time with their patients.
- Accurate and compliantNotes are precise, and patient data is handled in line with privacy and security standards.
Client voice
We had an incredible experience working with this team. They built a complete React.js platform with a Prisma backend, integrating AI-powered voice-to-text and Stripe for payments. They handled the project end-to-end, delivering a product that was polished, functional, and exactly what we needed. From the start, they impressed us with their professionalism and collaborative approach. They didn’t just build to spec—they actively worked with us to refine our UI/UX, suggesting thoughtful improvements that made the platform smoother and more intuitive. They were also flexible and open to feedback, making changes quickly and effectively whenever we had suggestions or ideas. Their communication style made everything easier; they kept us updated throughout and were always available to answer questions or walk us through solutions. The team’s technical skills were clear in how they brought our vision to life. The AI voice-to-text integration works perfectly, and the Stripe payment setup is both secure and user-friendly, giving our customers a smooth experience. They even provided a helpful walkthrough of the platform at the end, making sure we understood how to use and manage everything on our own. Ultimately, they delivered a high-quality, professional product that’s already had a positive impact on our business. If you’re looking for a skilled, reliable, and responsive team, we highly recommend them. This was one of the best experiences we’ve had on Upwork, and we look forward to working with them again!
Chaithra KumarCo-founder, Australian healthtech startupThe stack
Deepgram powers live transcription and OpenAI generates the clinical notes. The front end is React with Tailwind CSS on a Node.js and PostgreSQL backend, with Python for data processing. Payments run through Stripe, and the platform is hosted securely on AWS.
OpenAI
Deepgram
React
Node.js
Python
PostgreSQL
AWS
Stripe
- Case study
High-performance multi-vendor marketplace
A three-sided marketplace (admin, vendor and customer applications on one backend) for a fast-moving e-commerce startup.Read the case study - Case study
Enterprise trainee assessment and management system
An exam, learning and certification platform that has grown with its client for over two and a half years and now serves more than 3,000 trainees.Read the case study






