Follow along on YouTube · Instagram · X → @aiwithrav
Merged code into libraries the AI/ML world actually uses.
| Repo | Contribution | Status |
|---|---|---|
joblib/joblib 4.1k★ |
Bug fix: accept any os.PathLike in dump()/load() (PR #1812) |
✅ Merged |
joblib/joblib 4.1k★ |
Docstrings for time-format helpers (PR #1811) | ✅ Merged |
huggingface/sentence-transformers 18.9k★ |
Regression tests for a core utility (PR #3855) | ✅ Merged |
huggingface/sentence-transformers 18.9k★ |
Documented the sparse-tensor → SciPy COO conversion (PR #3843) | ✅ Merged |
nltk/nltk 14.7k★ |
Regression tests for transitive_closure — cycles, reflexive closure, and input-mutation safety (PR #3703) |
✅ Merged |
py-pdf/pypdf 10.1k★ |
Bug fix: low-bit /DeviceRGB images decoded as RGB instead of palette — broken image extraction (PR #3929) |
✅ Merged |
py-pdf/pypdf 10.1k★ |
Bug fix: low-bit expansion never ran for unfiltered/inline images, raising unrecognized image mode (PR #3938) |
✅ Merged |
| python-jose · spaCy · HF Evaluate · aiokafka | Security fix & bug fixes | 🔄 In review |
7 merged across 4 libraries the AI/ML world runs on — more PRs landing over time.
18+ AI agents coordinate dynamically to generate personalised structured intelligence reports across Vedic Astrology, Numerology, Palmistry, Tarot, and Vastu Shastra — in 23 Indian languages.
Enterprise-grade AI system that maps bench employees to open roles using 5-layer Hybrid RAG — with L1+L2 semantic cache, SSE streaming, circuit breaker, and LLM-as-judge guardrails. SonarQube Quality Gate PASSED.
Enterprise IT incident response engine — zero vectors, zero embeddings, zero hallucinated commands. Every kubectl command is pulled verbatim from the database. Three completely separate knowledge panels ranked by priority for every incident.
A LangGraph pipeline where 5 specialist agents collaborate to generate and optimise marketing campaigns — with a learning engine that uses past performance data to improve the next run automatically. ROI improves run over run.
An agent platform where any application can plug in an AI agent with just a YAML file — persona, tools, behaviour, all configured at runtime. Per-agent lock/unlock for token protection. Works with any model.


