Leveraging AI, IoT, quantum tech to solve real-world problems in personal and industrial domains.
View ProjectsHi, I'm Mohammad Tihame — Founder & Technical Visionary at MAEBR, and a B.Tech scholar specializing in Artificial Intelligence. I build intelligent, scalable systems that bridge research and real-world impact.
I architect and ship end-to-end AI and software solutions — from NLP pipelines and computer vision models to full-stack web platforms. My work spans the complete stack: Python-driven ML/DL backends, React and .NET frontends, IoT and edge deployments, and production-grade infrastructure. I don't just prototype — I ship.
C++, Python (NumPy, Pandas, PyTorch), React, .NET, FastAPI, IoT/Hardware-AI integration, ML/DL model development and optimization.
I operate at the intersection of intelligent automation and human-centered software — through MAEBR, active open-source contributions, and freelance AI/ML engineering. I'm an IEEE member with a commitment to high-impact, production-ready work. My goal is straightforward: solve hard problems, build things that last, and leave every system better than I found it.
Building intelligent systems that learn and adapt.
Connecting the physical and digital worlds.
From smart homes to industrial automation.
Next-gen security with quantum AI.
pushing boundaries
AI for all
from prototype to production
responsible by design
community first
Two Production-Grade Open Source Projects demonstrating end-to-end engineering across system design, algorithms, protocols, and AI integration.
High-Performance C++ Search Engine · MCP Server · RAG Pipeline
SearAG is a production-grade search engine written in modern C++ (C++20/23) that combines traditional information retrieval with modern AI. It exposes capabilities through the Model Context Protocol (MCP) for AI agent integration, and supports Retrieval-Augmented Generation (RAG) with pluggable LLM backends (Claude, Gemini, OpenAI, local models). Features include Hybrid Search (BM25 + semantic vector search with RRF), blazing-fast memory-mapped indexes, and SIMD vector ops with sub-100ms query latency on 1M-document corpora.
Fully Local RAG · LangGraph Agentic AI · Safe Computer Control · Python
A fully local Retrieval-Augmented Generation agent that combines Ollama for local LLM inference, ChromaDB as a vector database, LangGraph for ReAct-style agentic orchestration, and the Model Context Protocol (MCP) Python SDK for modular tool integration. All processing runs on-device — no data is sent to external APIs. It includes a unique Safe Computer Control module allowing AI-assisted system actions with mandatory human approval before any execution, enforcing a strict command allowlist and audit logging.
Professional PACS-Style App · Deep Learning · Real-time Analysis
A professional PACS-style medical imaging application that uses deep learning to analyze chest CT scans and detect potential malignancies. The platform features 4-class classification (Adenocarcinoma, Large Cell, Normal, Squamous Cell), real-time instant AI-powered insights upon scan upload, and color-coded confidence bars. It sports a modern glassmorphism design with a collapsible sidebar, floating controls, and thumbnail navigation.