Skip to content

muend/muend

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

20 Commits
 
 
 
 

Repository files navigation

Muhammed Enes Duran — GeoAI · Spatial Data Science · Applied Simulation Systems · Remote Sensing ML

Portfolio Agri-DSS live app FOUNDER.EXE on itch.io Email

I build production-grade GeoAI, spatial decision-support systems, and applied simulation products: ArcGIS automation exposed to LLM agents, deep-learning pipelines for satellite imagery, browser-native DSS tools, and realistic strategy simulations that turn complex real-world rules into usable software.

100 MCP geoprocessing tools 147 neighborhoods modeled Published game: FOUNDER.EXE Live SaaS product TÜBİTAK 2209-A grant funded


Currently Shipping

Product Type Status Link
arcgis-mcp-bridge GeoAI infrastructure / MCP tooling Active · PyPI · Glama A-rated Repository
agri-dss Spatial Decision Support System Live tarimsalkoridor.online
FOUNDER.EXE Browser-based startup simulation Published Play on itch.io
agri-unet Remote sensing deep learning research Funded research Repository

Product / Infrastructure / Research Map

Track What I build Representative work
Product Systems Browser-native tools, simulations, public-facing decision-support products agri-dss, FOUNDER.EXE
GeoAI Infrastructure Secure automation layers that connect GIS runtimes, LLM hosts, and geoprocessing workflows arcgis-mcp-bridge
Research & Modeling Reproducible spatial analysis, remote-sensing ML, urban resilience and econometric studies agri-unet, kutri-resilience-index, turkiye-housing-prices-pandemic

Technical Core & Methodology

I focus on structural data science, spatial machine learning, and applied product systems. My work connects rigorous mathematical models with interfaces people can actually use: from GIS automation layers and satellite-imagery pipelines to static, browser-native decision-support and simulation products.

Core AI / ML
Python PyTorch TensorFlow scikit-learn NumPy pandas SciPy

Spatial Data Science
GeoPandas Shapely ArcPy Rasterio PySAL

Systems, MLOps & Product Engineering
FastAPI Pydantic Docker Pytest Ruff Mypy JavaScript HTML5 CSS3 Git


Flagship Systems & Product Work

An institutional-grade Model Context Protocol (MCP) framework exposing exactly 100 specialized geoprocessing tools directly to LLM hosts and intelligent agents — turning ArcGIS Pro into a programmable backend for AI workflows.

  • Process Isolation: Strictly decoupled multi-process architecture: async MCP server core and isolated ArcPy worker subprocess.
  • Security Layer: PathGuard sandbox with prefix validation over filesystem and geodatabase paths before algorithmic execution.
  • Distribution & Verification: Published on PyPI, listed on Glama with A-rated MCP quality, and validated through 81 passing tests with mocked ArcPy.
  • Agentic GIS: Designed for local, controlled ArcGIS Pro automation from LLM hosts without collapsing the LLM runtime into the licensed GIS interpreter.
flowchart LR
    A["LLM Host / AI Agent"] -->|MCP protocol| B["MCP Server<br/>Async Core"]
    B --> C{"PathGuard<br/>sandbox"}
    C -->|path validated| D["Isolated Worker<br/>Subprocess"]
    C -.->|rejected| X["Blocked"]
    D --> E["ArcGIS Pro · ArcPy"]
    E --> F["100 geoprocessing tools"]
Loading

2. agri-dss  —  Live at tarimsalkoridor.online

A fully client-side Spatial Decision Support System for the Western Antalya agricultural corridor — 5 districts, 147 neighborhoods (Demre, Finike, Kaş, Kemer, Kumluca) — turning local agronomic and economic knowledge into a concrete, printable plan for each neighborhood.

  • Zero-Backend Static Architecture: Vanilla JS application against a decoupled data.json layer — no backend, no build step, trivially hostable and auditable.
  • DRY Data Contract: Compact cropSets / longTermCrops / regions schema resolved at runtime; recommendations can be updated by editing data only.
  • Swiss / Typographic Interface: Guided stepper, corridor diagram, live counters, and clean A4 print output for village boards and cooperatives.

3. FOUNDER.EXE  —  Browser Game / Startup Simulation

A browser-based startup simulation game modeling the practical stress of founding a company in Türkiye or the USA: taxes, incorporation choices, grants, investor expectations, cash flow, regulatory friction, AI-assisted advising, and bankruptcy/exit outcomes.

  • Rules-Driven Simulation: Country-specific rule sets for Türkiye and the USA, including tax filings, company types, grants, regulatory constraints, inflation/currency pressure, investor interactions, and runway management.
  • Idea-Aware Onboarding: The player writes a startup idea; the system analyzes sector, regulatory requirements, global potential, and market path before shaping the simulation.
  • Optional Live AI Layer: Gemini, OpenAI, or Anthropic can be connected with the player's own API key; the game remains playable with scripted fallbacks when offline.
  • Browser-Native Product: Built as a static web simulation with local save persistence, rule-based economic modeling, country-specific startup/legal parameters, and optional LLM-powered advisory flows.

Play FOUNDER.EXE on itch.io


Academic & Research Projects

Reproducible, peer-review–oriented studies in spatial econometrics, urban resilience, and deep learning for remote sensing.

agri-unet  ·  Deep Learning / CV  ·  TÜBİTAK 2209-A

The codebase for my TÜBİTAK 2209-A research project (University Students Research Projects Support Program) — an accepted, grant-funded study. A U-Net semantic segmentation pipeline for agricultural pattern identification from high-resolution, multi-temporal satellite imagery, extracting field parcels and crop structures for downstream suitability modeling.

turkiye-housing-prices-pandemic  ·  Spatial Econometrics

Region-level analysis of Türkiye's housing market that separates real (inflation-adjusted) price growth from inflation, comparing the six years before and after the COVID-19 pandemic (2014–2025).

  • Reproducible Python notebook with high-resolution choropleth figures and a House Price Index (HPI) deflation pipeline.
  • LISA (Local Indicators of Spatial Association) analysis to detect statistically significant regional clusters and spatial outliers.

kutri-resilience-index  ·  Composite Indicators

A reproducible urban-territorial resilience index prototype for Kaş / Bayındır, Antalya, based on a five-pillar composite indicator framework.

  • Transparent indicator normalization and weighting methodology with fully reproducible notebooks and figures.
  • Bridges quantitative spatial analysis with applied territorial planning.

Repository Map

Start here depending on what you want to inspect:

Repository / Product Best entry point Why it matters
arcgis-mcp-bridge GeoAI infrastructure ArcGIS Pro automation for LLM/MCP workflows with runtime isolation, path safety, PyPI distribution, and Glama A-rated MCP quality
agri-dss Product / DSS Live, zero-backend spatial decision-support system for agricultural planning
agri-unet Deep learning research U-Net remote-sensing segmentation pipeline for agricultural pattern identification
kutri-resilience-index Composite indicators Reproducible urban-territorial resilience index methodology
turkiye-housing-prices-pandemic Spatial econometrics Regional housing-price analysis with deflation and LISA workflows
FOUNDER.EXE Applied simulation product Published browser game modeling startup formation, taxes, funding and regulatory friction

Live Metrics

arcgis-mcp-bridge PyPI version arcgis-mcp-bridge monthly PyPI downloads arcgis-mcp-bridge GitHub stars Glama A-rated MCP quality Profile views


Product Notes / Case-Study Hooks

These are the product narratives I am currently expanding into separate case studies:

  • FOUNDER.EXE: encoding company formation, tax pressure, grants, investment logic, regulatory constraints, and optional AI advising into a playable browser simulation.
  • Agri-DSS: compressing agricultural suitability and local economic knowledge into printable, village-level decision plans.
  • arcgis-mcp-bridge: separating licensed GIS execution from AI host runtimes while preserving secure, discoverable, and quality-scored geoprocessing access.

Active Research & Product Workspace

  • Computer Vision for Remote Sensing: Automated pipelines for agricultural pattern identification and urban object extraction from high-resolution multi-temporal satellite imagery using CNN and U-Net segmentation models.
  • Agentic GIS & MCP Tooling: Secure local automation layers that expose GIS operations to LLM agents without compromising ArcGIS Pro runtime isolation.
  • Applied Simulation Systems: Browser-native simulations that encode real-world regulatory, financial, and decision processes into interactive products.
  • Urban Resilience Forecasting: Predictive spatial suitability matrices and long-term territorial resilience frameworks using robust statistical models.

Focus

Open to collaborative tracks involving production-grade Data Science, Spatial Machine Learning pipelines, automated GeoAI systems architecture, and applied simulation products.

About

Data Scientist & ML Engineer specializing in Spatial Data Science, GeoAI infrastructure, computer vision for remote sensing, and automated MLOps pipelines.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors