{"id":34532852,"url":"https://github.com/djleamen/troubleshoot-agent","last_synced_at":"2026-05-28T15:31:35.742Z","repository":{"id":325377817,"uuid":"1100165068","full_name":"djleamen/troubleshoot-agent","owner":"djleamen","description":"Building Intelligent Troubleshooting Agents by Microsoft on Coursera","archived":false,"fork":false,"pushed_at":"2026-02-14T06:36:55.000Z","size":252,"stargazers_count":1,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-02-14T14:31:14.352Z","etag":null,"topics":["ai","bert","classification","coursera","coursework","decision-trees","feature-selection","fine-tuning","lora","ml","nlp","pruning","python","pytorch","qlora","quantization","scikit-learn","sentiment-analysis","study","tensorflow"],"latest_commit_sha":null,"homepage":"https://www.coursera.org/account/accomplishments/verify/WYQWITZGLVKC","language":"Jupyter Notebook","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/djleamen.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":"2025-11-19T23:28:08.000Z","updated_at":"2026-02-14T06:36:58.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/djleamen/troubleshoot-agent","commit_stats":null,"previous_names":["djleamen/troubleshoot-agent"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/djleamen/troubleshoot-agent","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/djleamen%2Ftroubleshoot-agent","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/djleamen%2Ftroubleshoot-agent/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/djleamen%2Ftroubleshoot-agent/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/djleamen%2Ftroubleshoot-agent/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/djleamen","download_url":"https://codeload.github.com/djleamen/troubleshoot-agent/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/djleamen%2Ftroubleshoot-agent/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":33615489,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-26T15:22:16.424Z","status":"online","status_checked_at":"2026-05-28T02:00:06.440Z","response_time":99,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"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":["ai","bert","classification","coursera","coursework","decision-trees","feature-selection","fine-tuning","lora","ml","nlp","pruning","python","pytorch","qlora","quantization","scikit-learn","sentiment-analysis","study","tensorflow"],"created_at":"2025-12-24T05:53:08.032Z","updated_at":"2026-05-28T15:31:35.736Z","avatar_url":"https://github.com/djleamen.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Building Intelligent Troubleshooting Agents\n\nThis repository contains coursework and practice activities from the **Building Intelligent Troubleshooting Agents** course by Microsoft on Coursera. The course covers the development of AI-powered agents capable of diagnosing, analyzing, and resolving technical issues using machine learning and natural language processing.\n\n## Course Structure\n\n### 1. LLM Fine-tuning\nLearn how to fine-tune large language models for troubleshooting applications.\n\n**Activities:**\n- **Model \u0026 Dataset** (`model\u0026dataset.*`): Introduction to selecting appropriate models and datasets for fine-tuning\n- **Prepare Data** (`prepdata.*`): Data preprocessing and preparation techniques for LLM training\n- **Fine-tune LLM** (`finetunellm.*`): Step-by-step guide to fine-tuning language models\n- **LoRA** (`lora.ipynb`, `lora_example.py`): Low-Rank Adaptation techniques for efficient fine-tuning\n- **QLoRA** (`qlora.*`): Quantized Low-Rank Adaptation for memory-efficient training\n- **PEFT** (`peft.ipynb`, `peft_example.py`): Parameter-Efficient Fine-Tuning methods\n- **LLM Tuning** (`llmtuning.*`): Advanced tuning strategies and hyperparameter optimization\n- **Compare Fine-tuning** (`comparefinetuning.*`): Comparative analysis of different fine-tuning approaches\n- **Apply Evaluation Methods** (`applyevalmethods.*`): Techniques for evaluating fine-tuned models\n\n### 2. Fundamentals of AI Agents\nUnderstand the core concepts and requirements for building AI troubleshooting agents.\n\n**Activities:**\n- **Requirements** (`requirements.*`): Identifying and defining requirements for troubleshooting agents\n- **Troubleshooting Agent** (`troubleshootagent.*`): Building a basic troubleshooting agent from scratch\n- **Knowledge Base** (`data/troubleshooting_knowledge_base.json`): Creating and structuring knowledge bases for agent reference\n\n### 3. NLP (Natural Language Processing)\nIntegrate NLP techniques to enable agents to understand and process user queries.\n\n**Activities:**\n- **NLP for Troubleshooting** (`nlpfortroubleshoot.*`): Applying NLP techniques to troubleshooting scenarios\n- **Sentiment Analysis** (`sentimentanalysis.*`): Analyzing user sentiment to prioritize and understand issues\n- **Integrating NLP** (`integratingnlp.*`): Combining NLP with troubleshooting logic\n- **Agent Interface** (`Agent Interface/`): Building a web-based chatbot interface for user interaction\n  - `index.html`: Front-end interface\n  - `chatbot.js`: Chatbot logic and interaction handling\n\n### 4. Troubleshooting Agent\nDevelop comprehensive troubleshooting agents with classification, decision-making, and solution recommendation capabilities.\n\n**Activities:**\n- **Classification Models** (`classificationmodels.*`): Training models to classify different types of issues\n- **Evaluation of Classification** (`evalclassification.*`): Assessing the performance of classification models\n- **Mechanisms** (`mechanisms.*`): Implementing error handling and validation mechanisms\n- **Logging** (`logging.ipynb`, `logging_example.py`): Setting up logging for debugging and monitoring\n- **Decision Making** (`decisionmaking.*`): Building decision trees and logic for root cause analysis\n- **Solution Recommendation** (`solutionrec.*`): Creating recommendation systems using KNN and other algorithms\n- **Implementation** (`implementation.*`): Integrating all components into a complete troubleshooting agent\n- **Troubleshooting Agent** (`troubleshootagent.*`): Full implementation with anomaly detection, root cause analysis, and solution recommendation\n\n### 5. Testing\nTest, evaluate, and optimize troubleshooting agents for production deployment.\n\n**Activities:**\n- **Test Cases** (`testcases.*`): Designing comprehensive test cases for ML systems\n  - Typical case testing\n  - Edge case handling\n  - Error handling with missing values\n- **Evaluate Effectiveness** (`evaleffectiveness.*`): Measuring agent performance using:\n  - Accuracy and precision metrics\n  - Response time analysis\n  - Resource usage monitoring (CPU/Memory)\n  - Stress testing\n  - Cross-validation\n- **Optimization** (`optimization.*`): Implementing optimization techniques:\n  - Model pruning\n  - Quantization with TensorFlow Lite\n  - Feature selection\n- **Test \u0026 Optimize** (`test\u0026optimize.*`): Combined testing and optimization workflow\n\n## Key Technologies\n\n- **Machine Learning**: scikit-learn, TensorFlow, PyTorch\n- **NLP**: Transformers, BERT, Sentiment Analysis\n- **Fine-tuning**: LoRA, QLoRA, PEFT\n- **Optimization**: TensorFlow Model Optimization, Pruning, Quantization\n- **Testing**: pytest, ipytest\n- **Logging**: Python logging module\n\n## Outputs\n\n- **Fine-tuned Models**: `fine_tuned_bert/` - BERT model fine-tuned for sentiment analysis\n- **Logs**: `ml_errors.log`, `ml_pipeline.log` - Error and pipeline logging outputs\n- **Results**: Checkpoints and training results from various experiments\n\n---\n\n*Course by Microsoft on Coursera*","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdjleamen%2Ftroubleshoot-agent","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fdjleamen%2Ftroubleshoot-agent","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdjleamen%2Ftroubleshoot-agent/lists"}