Koshika

Koshika is an offline-first, open-source mobile app that helps people understand their lab reports without depending on the cloud. Most blood test reports are delivered as PDFs filled with abbreviations, reference ranges, and clinical terms that are difficult for most people to interpret. Even when the information is important, it is usually trapped inside a static document.

On top of that, many existing health apps require users to upload sensitive medical data to remote servers before they become useful.

Koshika is built around a different idea: your health data should stay on your device and still be useful, understandable, and searchable even without internet access. The app imports PDF lab reports directly on the phone, extracts biomarker values locally, and turns them into a structured record that users can actually explore over time. Instead of just storing reports,

Koshika also supports FHIR R4 export, making health records more interoperable and portable instead of locking them inside one app. For FOSS Hack 2026, the goal is to build something genuinely useful: not a generic dashboard, but a practical, privacy-respecting health tool that solves a real problem for real people.

Features

- Offline-first mobile experience with no cloud dependency

- Local PDF lab report import and parsing

- Support for multiple Indian lab report formats

- OCR fallback for difficult or image-based PDFs

- Tracking for 63 biomarkers across multiple categories

- Trend charts, reference ranges, and borderline detection

- On-device AI chat grounded in the user’s own lab data

- Safety-aware response pipeline with validation and guardrails

- FHIR R4 export for structured, standards-based health data

- No accounts, no telemetry, and no mandatory internet connection

Tech Stack

- Framework: Flutter (Dart)

- Local storage: ObjectBox

- PDF extraction: syncfusion_flutter_pdf

- OCR: Google ML Kit + pdfx

- Charts and visualisation: fl_chart

- On-device LLM runtime: llamadart / llama.cpp with GGUF models

- Embeddings and semantic search: bge-small-en-v1.5 + HNSW vector indexing

- Health data export: fhir_r4

Why this project matters

Koshika is an attempt to show that health software can be open-source, useful, privacy-respecting, and offline-capable at the same time. If a person’s body generated the data, they should be able to access it, understand it, and keep control of it without having to hand it over to a cloud service first.