autollecto automates product cataloging using computer vision and local AI. Point a camera at a product, and the system identifies it, generates a structured description, and adds it to your catalog. Runs locally with Ollama or connects to cloud vision APIs.

Features

photo_camera

Vision pipeline

YOLO11n object detection + vision LLM for automatic product identification and description.

data_object

Structured output

Consistent JSON catalog entries via OpenAI SDK with schema validation.

cloud_off

Dual mode

Run fully local with Ollama or connect to cloud APIs. Same pipeline, your choice.

api

API-ready

Built for integration. Sprint 2 adds HTTP endpoints on top of the rebuilt core.

Philosophy

Manual product cataloging is slow, inconsistent, and doesn't scale. autollecto replaces human data entry with a vision pipeline: YOLO11 detects objects in frame, a vision LLM describes them, and structured output ensures every catalog entry follows the same schema. The entire system can run offline with Ollama.

Stack

Python with OpenAI SDK for unified LLM access (Ollama and cloud providers). YOLO11n for real-time object detection. Structured output with Pydantic validation. Designed for integration into larger inventory management systems.

Tech Stack

Python OpenAI SDK YOLO11 Ollama

Autollecto

Features

Philosophy

Stack

Tech Stack