v1.0.0 — Main Launch

First stable release of LAiSER for extraction and alignment workflows.

v0.5.0 — Knowledge & Task Pipeline

What's New

Knowledge & Task Pipeline

• Add Knowledge and Task FAISS indexes alongside existing Skills index
• Add prebuilt taxonomy indexes for Knowledge and Tasks
• Add pipeline build scripts for Knowledge and Task indexes

KST Extraction API

• Generalize AlignmentService to support all three index types (Knowledge,
  Skills, Tasks)
• Add concept extraction API for Knowledge, Skills, Tasks

Other

• Update Gemini integration
• Add Task Statements workbook

v0.4.1

Add source_url to taxonomy alignment output

V0.4.0.1 Streamlined Index Pipeline and Multi-Cloud LLM Support

Release Notes

Highlights

This release introduces a streamlined index pipeline and adds multi-cloud LLM support.

What's New

Streamlined Index Pipeline

The index pipeline has been simplified to make extraction and indexing workflows more efficient and easier to manage.

Multi-Cloud LLM Support

Support for multi-cloud LLM integration has been added, enabling the module to work across different cloud-based model providers.

Release v0.3.2: Architectural Refactor, Enhanced Platform Support, and Bug Fixes

🚀 Development Update

✨ Key Features & Improvements

• Architectural Refactoring: Implemented layered architecture with improved separation of concerns
• LLM Models: Added llm_models subpackage for better organization
• Documentation: Updated contributor guidelines with Conventional Commit standards

🐛 Bug Fixes

• Fixed dtype inconsistencies with GPTQ quantization, switched to AWQ quantization
• Resolved data_type error for Gemma2 models in vLLM, no longer supported
• Updated input exceptions to prevent unnecessary warnings
• Improved input data format to support user preferred labels
• Enhanced LLM response parsing

📚 Documentation Updates

• Added Semantic Versioning convention
• Updated README and contributor guides
• Added refactoring guide

🔧 Technical Changes

• Updated project configuration in pyproject.toml
• Enhanced main application entry point at __init__.py
• Improved error handling and data access patterns
• Implemented the Service layer to separate business logic

Release v0.3.0: LLM-Driven Skill Extraction & Taxonomy Mapping

LAiSER Extract Module v0.3 🚀

🚀 Development Update

✨ Key Features & Improvements

1. LLM-Driven Skill Extraction

• Replaces the old raw description embedding approach.
• Skills are now explicitly extracted from cleaned job descriptions using structured LLM prompts.

2. Preprocessing Layer Added

• Job descriptions are cleaned before extraction to remove irrelevant or misleading content (PII, HR/legal text, fluff, etc.).
• Ensures that alignment happens only on meaningful skill signals.

3. Testing Pipeline Introduced

• Multiple test files added to validate core package functionality.
• Strengthens reliability and long-term maintainability.

---

📂 Release Resources

• Source code: GitHub Repository
• Python package : PyPi

---

📌 Notes

This release marks the transition from a doc-embedding-only approach to a hybrid pipeline (Preprocess → LLM Extract → FAISS Align), producing cleaner, more interpretable, and auditable outputs.

Skill-Extraction v0.2.2 (ESCO + KSA)

Release v0.2.2 · Skill-Extraction Refactor (“ESCO + KSA”)

⚡PR⚡#101

✨ Highlights

1. Taxonomy-aware skill extraction
   • Integrates a FAISS index of the ESCO skills taxonomy.
   • Skill_Extractor.get_top_esco_skills() now returns {Skill, index, score} enabling deterministic Skill Tag values (ESCO.<index>).

2. KSA enrichment with vLLM
   • New helper get_ksa_details() generates Knowledge Required and Task Abilities lists for each skill.
   • Automatically invoked when a GPU/vLLM backend is available.

3. Unified output schema
   The extractor returns a tidy DataFrame with seven columns:
   Research ID, Description, Raw Skill, Knowledge Required, Task Abilities, Skill Tag, Correlation Coefficient.

---

🔧 Detailed Changes

Area	Description
utils.py	get_top_esco_skills() enhanced to include ESCO index and similarity score.
llm_methods.py	Added get_ksa_details() plus supporting imports.
skill_extractor.py	• Ensured self.index is always defined. • build_faiss_index_esco() / load_faiss_index_esco() now instance methods storing the index under laiser/input. • New taxonomy-first pipeline inserted at the top of extractor(); legacy alignment kept for fallback.

---

⚠️ Deprecated / To Be Removed

• align_skills() and align_KSAs() will be dropped in v0.3 once consumers migrate to the new output format.

---

🚧 Known Issues / Roadmap

1. JSON parsing in get_ksa_details() needs additional resilience checks.
2. LLM calls are still executed per skill; batching will come in v0.3.
3. Duplicate import json lines remain in llm_methods.py.
4. Consider CPU-only fallback for KSA generation.
5. Persistence of the ESCO vector index should move to a cloud vector DB.
6. vLLM isn't supported on MPS/MacOS as of now.

---

⬆️ Upgrade Notes

pip install -U laiser==0.2.2 

No changes to input parameters are required, but downstream code should read the new seven-column schema.

---

Next up

0.2 → 0.3

• adding batching and dropping deprecated APIs; increment patch for bug fixes.