{"id":47873350,"url":"https://github.com/scorpia2004/lazarus","last_synced_at":"2026-04-04T01:00:14.107Z","repository":{"id":348533905,"uuid":"1198537825","full_name":"SCORPIA2004/lazarus","owner":"SCORPIA2004","description":"poc for Erdemir","archived":false,"fork":false,"pushed_at":"2026-04-01T14:16:43.000Z","size":3,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-04-04T01:00:02.259Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":null,"has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/SCORPIA2004.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-04-01T14:15:05.000Z","updated_at":"2026-04-01T14:16:47.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/SCORPIA2004/lazarus","commit_stats":null,"previous_names":["scorpia2004/lazarus"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/SCORPIA2004/lazarus","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/SCORPIA2004%2Flazarus","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/SCORPIA2004%2Flazarus/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/SCORPIA2004%2Flazarus/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/SCORPIA2004%2Flazarus/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/SCORPIA2004","download_url":"https://codeload.github.com/SCORPIA2004/lazarus/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/SCORPIA2004%2Flazarus/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":31383636,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-04-03T23:20:52.058Z","status":"ssl_error","status_checked_at":"2026-04-03T23:20:51.675Z","response_time":107,"last_error":"SSL_read: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":[],"created_at":"2026-04-04T01:00:12.347Z","updated_at":"2026-04-04T01:00:14.056Z","avatar_url":"https://github.com/SCORPIA2004.png","language":null,"funding_links":[],"categories":[],"sub_categories":[],"readme":"# Idea Verification System — RAG Double-Checker for BERT Classifier\n\n\u003e **POC Project** | Erdemir | Timeline: 6-8 days\n\n---\n\n## Overview\n\nErdemir is a steel production company with ~10,000 employees. As part of a long-standing company tradition, every worker submits ideas for improving work processes and production. To date, over **300,000 ideas** have been collected.\n\nAn existing **BERT-based NLP classifier** has been developed to automatically evaluate incoming ideas and output an `Approve` / `Reject` decision, achieving **\u003e70% accuracy**. The goal of this project is to build a **RAG (Retrieval-Augmented Generation) system** that acts as a **secondary verification layer** on top of the BERT classifier.\n\n---\n\n## Problem Statement\n\nThe BERT classifier processes a plain-text idea submission and outputs a binary decision:\n\n```\nInput:  Plain-text idea (submitted by employee)\nOutput: Approve | Reject\n```\n\nWhile the BERT system performs well, a second-opinion layer is needed to increase confidence and reduce misclassifications. This project builds that layer as a RAG system.\n\n---\n\n## Solution Architecture\n\n```\nEmployee Idea (plain text)\n        │\n        ▼\n ┌─────────────┐\n │    BERT     │  ──── Approve / Reject\n │ Classifier  │\n └─────────────┘\n        │\n        ▼ (BERT output + original idea)\n ┌─────────────┐\n │  RAG System │  ──── Confirm / Override\n │ (this repo) │\n └─────────────┘\n        │\n        ▼\n  Final Decision\n```\n\nThe RAG system receives both the **original idea text** and the **BERT classification output**, retrieves relevant context from the historical idea dataset, and produces a verification judgment — either confirming or challenging the BERT result.\n\n---\n\n## Scope\n\nThis repository covers the **POC (Proof of Concept)** phase:\n\n- Set up and configure the RAG system for local machine deployment (RTX 5090 or equivalent, targeting models like QwQ 3.5 or similar)\n- Integrate with the existing BERT classifier output\n- Evaluate RAG verification accuracy against the labeled dataset\n- Determine whether fine-tuning is required based on data characteristics\n\n---\n\n## Current Status\n\n| Component | Status |\n|---|---|\n| BERT classifier | ✅ Exists, working |\n| Labeled idea dataset | ✅ Available |\n| RAG system | 🔧 In development (this repo) |\n| Fine-tuning | ❓ TBD — pending data review |\n\n---\n\n## Project Details\n\n- **Client:** Erdemir\n- **Scale:** 300,000+ historical ideas, 10,000 employees\n- **Deployment target:** Local machine (RTX 5090 / QwQ 3.5 or similar)\n- **POC estimate:** 6–8 person-days\n- **Full build timeline:** 5–6 months\n\n---\n\n## Action Items\n\n- [ ] Obtain project proposition and technical documentation from Eren\n- [ ] Review dataset to determine fine-tuning requirements\n- [ ] Set up RAG pipeline and connect to BERT output\n- [ ] Evaluate end-to-end system performance\n\n---\n\n## Open Questions\n\n- Is fine-tuning required for the RAG model? (Pending data review)\n- What retrieval strategy best fits the idea domain? (semantic search, keyword, hybrid?)\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fscorpia2004%2Flazarus","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fscorpia2004%2Flazarus","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fscorpia2004%2Flazarus/lists"}