TYPESCRIPTPYTHONSUPABASERAG

ENGRAM

B2B AI SaaS — LLM-powered enterprise memory.

ARCHITECTURE

Data Sources

Slack, Notion, Drive, Email

Ingestion Pipeline

Semantic chunking & embedding

Vector Store

Supabase pgvector

RAG Engine

Contextual retrieval + LLM

PRODUCT GALLERY

SCREENSHOT PLACEHOLDER
—
IMG 1 OF 3
SCREENSHOT PLACEHOLDER
—
IMG 2 OF 3
SCREENSHOT PLACEHOLDER
—
IMG 3 OF 3

MY ROLE

Co-Founder & Engineer

Led technical architecture decisions and full-stack implementation from day one.

Data Pipeline Design

Built semantic data pipelines that transform fragmented enterprise knowledge into unified, queryable vectors.

RAG Architecture

Architected retrieval-augmented generation search interfaces for contextual, accurate answers.

Multi-Source Integration

Designed the knowledge base integration layer supporting multiple enterprise data connectors.

TECH STACK

FRONTEND

  • Next.js 15
  • TypeScript
  • Tailwind CSS
  • Framer Motion

BACKEND

  • Supabase
  • PostgreSQL
  • pgvector
  • Edge Functions

AI / ML

  • OpenAI API
  • RAG Pipeline
  • Semantic Embeddings
  • Vector Search