Our Approach Technology Portable Record Connectivity Privacy AI Pipeline Get Started

Care without the paperwork.

ChartLite is a voice-powered documentation tool built for primary care. It operates 100% offline and is completely free, translating spoken consultations into structured clinical records.

A smiling healthcare professional holding a mobile device.

"A clinic's most valuable resource is the clinician's time. ChartLite restores that time by completely removing the friction of data entry."

As natural as a conversation.

We believe technology should adapt to how you practice medicine, not force you to change your workflow. By utilizing advanced on-device speech recognition, ChartLite listens to the natural dialogue between you and your patient.

There are no complex menus to navigate or checkboxes to fill out. You speak. The application understands.

  • 01

    Speak Your Assessment

    Discuss symptoms and treatment with your patient normally while the device records.

  • 02

    Instant Structure

    The system automatically extracts the diagnosis, vital signs, and prescribed medications.

  • 03

    Review and Save

    A formatted clinical note is generated for your brief review and immediate approval.

A clinician holding a phone showing the ChartLite voice-to-chart interface in a consultation room.

100% Offline. Scales as you grow.

Many systems require expensive servers and uninterrupted broadband. We designed ChartLite to function flawlessly even when those aren't guaranteed.

It operates entirely offline on standard Android devices, with battery-aware inference that queues transcripts and processes them in efficient batches. When reliable internet or advanced devices are available, ChartLite adds secure peer-to-peer sync — including cross-facility patient sharing — optional cloud AI, and enterprise-grade reporting integrations.

Whether you are operating a rural healthcare outpost or a fast-paced urban practice, the platform adapts to your infrastructure. No matter the setting, your data is protected with bank-standard encryption.

ChartLite is open-source and will always be completely free for clinical use. No subscriptions, no hidden fees.

  • 01 Zero ongoing licensing fees.
  • 02 Fully functional without internet.
  • 03 Unlocks optional cloud AI & reporting when connected.
  • 04 Built for both remote clinics & urban practices.

Enterprise-Grade Architecture.

Built on modern, verifiable technologies to ensure reliability, security, and interoperability in the most demanding environments.

On-Device AI Inference

Unlike traditional cloud-dependent EMRs, ChartLite executes entirely locally. It utilizes Meta Omnilingual ASR (1600+ languages via ONNX) for speech recognition and a quantized Qwen 3.5 model via llama.cpp for clinical extraction. An on-device RAG pipeline retrieves only the relevant diagnoses and drugs per transcript — keeping context windows lean and accurate on resource-constrained hardware.

Zero-Trust Security

Data at rest is protected by a SQLCipher encrypted database. In completely offline settings, vital synchronization is achieved using an AES-256-GCM encrypted binary payload over standard SMS networks. Play Integrity device attestation verifies app authenticity before any cloud API access.

Global Interoperability

Clinical data is stored natively for speed, with FHIR R4 bundles generated on-demand for external systems like DHIS2 and OpenMRS. A built-in billing engine automates ICD-10 to CPT/HCPCS mapping, E/M level coding, and SOAP note generation from structured encounter data. Ready for insurance claims and national health information exchanges across 8 countries.

ChartLite is in active development.

We're building in the open and iterating fast. The project is free and open-source — you're welcome to try it, explore the code, and see how it works.

If you run a clinic or health facility and want to explore a pilot, we'd love to hear from you. And whether you're a clinician, developer, or health systems researcher — your feedback helps us build something better.

View on GitHub Get in Touch