AI-Driven NoSQL Schema Mapping and Setup

2 min read

AI-Driven NoSQL Schema Mapping and Setup

  • Day: 2024-09-16
  • Time: 16:55 to 17:25
  • Project: Dev
  • Workspace: WP 2: Operational
  • Status: In Progress
  • Priority: MEDIUM
  • Assignee: Matías Nehuen Iglesias
  • Tags: AI, Nosql, Openai, Schema Mapping, Entity Extraction

Description

Session Goal

The primary goal of this session was to explore and establish a workflow for mapping entities from natural language texts into NoSQL schemas using AI agents.

Key Activities

  • Mapping Entities with AI Agents: Discussed strategies for utilizing AI agents within a NoSQL schema to map extracted entities from resolution texts.
  • Dynamic Workflow Development: Outlined a refined workflow for dynamically mapping extracted information from government resolution texts to a predefined NoSQL schema.
  • Text-to-SQL Evaluation: Analyzed the differences between Text-to-SQL tasks and the current goal of mapping legal texts into NoSQL schemas.
  • OpenAI Tools Utilization: Detailed the process of using OpenAI’s tools, particularly GPT-4, for extracting entities and converting them into NoSQL schemas.
  • Guide for AI Schema Generation: Provided a structured guide for generating accurate NoSQL schemas using OpenAI Functions and fine-tuning techniques.
  • Questions and Considerations: Raised specific questions about OpenAI Functions and JSON schema definition, focusing on error handling and API limitations.
  • Project Setup with Bash Commands: Offered a set of Bash commands for project setup, including preparing datasets and implementing OpenAI Functions.

Achievements

  • Established a comprehensive workflow for AI-driven NoSQL schema mapping.
  • Clarified the use of OpenAI tools and functions for schema generation.
  • Developed a project setup guide using Bash for efficient implementation.

Pending Tasks

  • Further exploration of API limitations and cost considerations for OpenAI Functions.
  • Addressing specific questions raised about JSON schema structure and error handling.

Evidence

  • source_file=2024-09-16.sessions.jsonl, line_number=5, event_count=0, session_id=c8f8f669da4538c0330a452191dd9ea7f6534bed753e0d33e813e4941610f26a
  • event_ids: []

Graph View

Backlinks