Havuz

Personal academic and research portfolio focused on AI, low-resource Urdu NLP, RAG systems, machine learning, and computer science.

Biologically-grounded adversarial training platform: cyclic Wake/Dream/Nightmare/Compress phases that accumulate model robustness without ca

🎙️ Real-Time Multi-Modal Speech Telemetry & Acoustic DSP Engine (FastAPI, Python 3.11, Autocorrelation Pitch, FFT Spectral Energy & Lexical

中国法律法规与案例结构化语料库：446 部法律/司法解释逐条 JSON（含效力状态）+ 278 指导性案例 + 445 公报案例，CC0

Native AI inference for PHP 8.3+ - run ONNX, GGUF (llama.cpp) and RubixML models directly in your PHP process via FFI. Chat, streaming, embe

Prepackaged medical pronunciation dictionary for voice AI TTS engines. 960+ drug names, clinical terms, anatomical terms, and medical acrony

First open-source AI knowledge layer for Togo - 62K+ documents, RAG API, fine-tuned LLM. Built for developers, startups and institutions i

A consistent, fair, benchmark for structured extraction.

Xing4.0-29B-Intent - schema-guided intent recognition and structured information extraction (intent and slot filling, entity, relation and e

Detect hallucinations in LLM responses. Verify every claim against source documents using hybrid STS + NLI. Works with LangChain, LlamaIndex

🔍 Detect AI-generated text and fingerprint which LLM wrote it. Open-source GPTZero alternative. Zero dependencies, works offline.

A high-accuracy, multilingual rule-based Sentence Boundary Detector (SBD) with a drop-in adapter for pysbd, delivering faster and more accur

LLM-powered conversation summarisation service for OCTO — turns group chats and threads into structured briefs with key decisions, open ques

Zero-cost epistemic uncertainty quantification & hallucination detection for LLMs (90,000x faster than Semantic Entropy)

AI-powered fake news detection system with multiple ML models and explainability.

SomNLP-Translate is an open, high-quality English–Somali parallel dataset designed to support research and development in machine translatio

Turn documents into high-quality instruction datasets with grounding, quality filtering, deduplication, and provenance.

A pure-Python structured memory benchmark for multi-agent LLM systems — context graph vs vector RAG vs raw history dump, five scenarios, 18

A free, privacy-first tool that flags the tells of AI-generated writing, overused vocabulary, rhetorical crutches, robotic sentence rhythm a

Train linear embedding adapters with triplet loss to align retrieval embeddings with your queries (RAG).

Python SDK for LLM guardrails with safety classification, PII detection, prompt injection defense, and grounding checks. Protect AI applicat

A modern, hands-on, self-paced course that takes you from your first line of Python to shipping a real AI-powered automation.

Locally fine-tuned Russian RAG models: document splitter, query expansion, retrieval embedder. Teacher distillation → GGUF → llama.cpp on AM

Better Estonian from your AI agent. Offline MCP server (no third-party calls) wrapping EstNLTK, EKI Reeglid + Riigi Teataja. 28 read-only to

AI-detector auditing toolkit: measures how often a detector flags genuine human writing, how stable its verdict is across seeds, and whether

LLM-free, layout-aware PDF chunking for RAG pipelines. Preserves document structure via regex and font heuristics. Written in pure Rust.

A structured knowledge base for large language model technologies, covering LLMs, Agents, RAG, model training, evaluation methodologies, and

A zero-allocation, header-only C++ BPE tokenizer for Qwen, built for maximum inference throughput.

Runs Hugging Face ONNX models on any Nx backend, with no per-architecture code

This repository is a comprehensive, hands on roadmap to Python Mastery and AI Engineering. It is designed to take a learner from the very fo

Privacy and security guardrails for Spring AI: request-scoped PII tokenization, least-privilege disclosure to tools, and tool authorization

Discover the citing papers that matter most — ranked by impact, relevance, and influence

Asking for What Was Never Requested: horizontal and vertical proactivity in LLM agents. Need-graph metrics (no LLM judge) and Q&D, which tra

Wenshu（文枢）— 面向人文社科研究的 AI 知识处理工作流：本地知识库 / RAG / 中文引文 / 理论谱系。AI knowledge workflow for humanities & social sciences: local knowledge base, RAG

面向小团队的一站式 NLP 小模型平台:数据集版本管理 · 训练 · 评估 · 在线部署 · Badcase 闭环,内置大模型 Prompt 盲测评测(人工 + AI 双指标)。

A free, open GenAI curriculum built from scratch. Start from zero, build real projects, and work your way up to production AI systems.

日本語フリガナ (ルビ) を扱う Rust 製ライブラリ + ローカル HTTP サーバー。ルールはすべてデータ駆動 (TOML)。

Remove AI text watermarks (SynthID-Text, KGW, Unigram) by regenerating text from its meaning — scheme-blind, runs on local models, every cla

Caveman prompting, measured. A two-channel evaluation framework for prompt compression and LLM token efficiency: what input vs output compre

Free, local-first academic integrity checker and MCP server: citation hallucination detection, AI-text and plagiarism signals, venue guideli

工厂本体驱动的数据问答框架（开源，Apache-2.0）：把中小企业台账/MES/ERP 变成可自然语言问答的语义知识图谱。CSV进答案出，确定性规则引擎+本体GraphRAG+向量混合检索，本地化数据不出厂，换行业只换词典。

RAGless is a semantic retrieval system that answers questions about your documentation, without using an LLM at runtime.

Train JEV-style System 1 decision models on Qwen3.5 and Qwen3.8-27B: 24-slot decision head, KL distillation, calibration and serving. 在 Qwen

Advanced RAG (Retrieval-Augmented Generation) system for private document intelligence. Uses OpenAI GPT-4o and Vector Databases to provide z

Salience-aware lexical trie for long-context compression.

A Retrieval-Augmented Generation (RAG) application for querying legal documents. It uses PostgreSQL, Elasticsearch, and LLM to provide summa

Open benchmark for expert collapse and routing efficiency in sparse MoE LLMs

Let your users ask your database anything, in their own words — safely. The model fills in a validated Query AST, never a query string, and

Accepted at ISCSLP 2026 · Target-level benchmark for raw-input Chinese news TTS pronunciation

Python RAG-based chatbot for document Q&A using retrieval-augmented generation