NS
Nishant OS
Initializing Nishant AI
LOADING0%
All work
ProductionLocal-First Knowledge Management · Feb 2026 – May 2026

ClarityStack

Turn long-form AI chat histories into structured, version-controlled knowledge.

0
Microservices
SRS · editor · ThreatLens
0
Client platforms
Next.js web · Android
Immutable
Ground truth
raw transcripts
SRS
Retention
spaced repetition

ClarityStack is an end-to-end, local-first knowledge-management system that extracts long-form AI chat histories into structured, version-controlled knowledge bases — while preserving the raw transcripts as immutable ground truth. A distributed FastAPI + SQLite backend powers dedicated services for spaced-repetition learning, document editing, and analytics, served to a Next.js web app and a native Android client.

The problem

The knowledge you build up inside AI chats is trapped in long, ephemeral transcripts. It's hard to revisit, impossible to version, and easy to lose — and naive summaries throw away the original ground truth.

The solution

ClarityStack extracts transcripts into structured, version-controlled knowledge bases while keeping the raw text immutable. Microservices handle spaced-repetition (SRS), rich-text/markdown editing, and specialized analytics (ThreatLens); markdown workflows feed directly into SRS and temporal tracking cards for long-term retention across web and Android.

System Architecture
AI Chat HistoriesExtractorstructureVersioned KB+ immutable rawSRS ServiceEditor ServiceThreatLensanalyticsWeb + AndroidNext.js · Compose
Challenges solved
  • 1Extracting structure without ever mutating the ground-truth transcript.
  • 2A distributed microservice backend that stays coherent across services.
  • 3One seamless experience across a Next.js web app and a native Android client.
Highlights
  • Local-first: raw transcripts stay as immutable ground truth.
  • Distributed FastAPI microservices for SRS, editing, and analytics.
  • Cross-platform Next.js web app + native Android (Jetpack Compose).
Technology
PythonFastAPISQLiteNext.jsTypeScriptJetpack ComposeMarkdownRAG
Next project
Diabetic Retinopathy Detection
Medical Computer Vision · Deep Learning