Limbi — Omni-Agent Orchestration Platform

A production-grade multi-agent orchestration platform with 89 specialized agents and 435 available actions. One command routes across engineering, security, cloud, DevOps, finance, and domain agents. Supports 19 LLM provider modes — local (Ollama, LM Studio, vLLM) and cloud (OpenAI, Anthropic, Groq, OpenRouter). Features graph-backed session memory, RAG pipelines, FastAPI backend, MCP server, and VS Code extension.

Multi-AI-Models Chat Support System

A secure platform orchestrating parallel responses from multiple LLMs simultaneously. Unified API abstraction across Groq, OpenRouter, Bytez, and Chutes with Tavily-powered real-time web research.

Hike.ai — Unified AI Orchestration Platform

Five production AI modules under one roof: News Flow with semantic search, multi-model Debate Arena, Regret AI for decision analysis, and Empathy AI — secured with Google OAuth 2.0 and bcrypt auth.

Local Knowledge Chatbot — RAG System

Fully offline RAG chatbot running entirely on-device. Zero external API dependencies. Automated scraping pipelines feed ChromaDB; Ollama serves local LLMs (Mistral, LLaMA). One-command Docker deployment.

Sales Forecasting ML Pipeline

End-to-end ML pipeline reading directly from structured databases — no manual exports. Automated feature engineering and model evaluation loops benchmark regression and tree-based models on sales time-series data.

NLP Transformation Engine

Advanced NLP pipeline that converts AI-generated text into natural, human-like writing. Configurable humanization, summarization, and readability optimization with real-time AI pattern detection across 18 signal types. Style-adaptive output targeting Casual, Professional, Academic, and Concise registers.