Open to onsite & remote roles — Dhaka, Bangladesh (UTC+6)

Md Mobinur Rahman

Full stack developer building early-warning and geospatial systems

Ten years writing the software behind national flood forecasts, cyclone alerts and farm advisories — the systems that reach people before the weather does. Web, mobile, GIS, and the machine learning in between.

[email protected] +880 1711 415554 LinkedIn GitHub
10+Years building production software
8National platforms and public apps shipped
5Government ministries served
9South Asian countries covered by one forecast platform

What I actually do

Most of my work sits at the point where meteorological data becomes a decision. A model output arrives every six hours; somewhere downstream a district officer needs a map, a farmer needs an advisory in Bangla, and a warning centre needs a bulletin ready to send. I build the pipelines, APIs, maps and apps that carry it across.

Day to day that means Python and Django on the backend, React or Angular on the front, Flutter for mobile, and PostGIS underneath almost everything. It also means Kafka queues, Docker images, Nginx configs and the 2 a.m. production call during a cyclone, because these systems are judged on the days they are needed most.

The AI side grew out of the same problem. Crop disease detection from a farmer's photo, landslide susceptibility from terrain rasters, forecast post-processing — computer vision and modelling used where they measurably shorten the path from data to action.

How I work

  • ●End to end: architecture, build, deploy, CI/CD, and the production support afterwards.
  • ●Comfortable across multidisciplinary teams — meteorologists, GIS specialists, agronomists, designers, QA.
  • ●Built for constraint: low bandwidth, older Android devices, users who need it to work offline and in Bangla.
  • ●Remote-ready across time zones, with a decade of distributed and multi-country project delivery.

Experience

Jan 2020 — presentDhaka, Bangladesh

Senior Software Developer

Regional Integrated Multi-Hazard Early Warning System (RIMES)

  • Architect and build web, mobile, GIS and AI applications for multi-hazard early warning, forecasting and decision support.
  • Backend services and REST APIs in Python (Django, FastAPI), TypeScript (NestJS) and PHP (Laravel).
  • Frontends in React, Next.js, Angular and TypeScript.
  • Geospatial work with Leaflet, Mapbox, ArcGIS, QGIS, GDAL and PostGIS.
  • Machine learning and computer vision with YOLOv8, TensorFlow, OpenCV, Scikit-learn, plus LLM and LangChain work for advisory tools.
  • AI for crop disease and pest detection, landslide susceptibility mapping and forecast-driven advisories.
  • Cross-platform mobile apps in Flutter, alongside native Android and iOS work.
  • Event-driven and microservice architectures using Apache Kafka and RabbitMQ.
  • Containerised deployment with Docker and Kubernetes on Linux infrastructure.
  • CI/CD pipelines and cloud deployment across AWS, Azure and GCP.
  • Database design across PostgreSQL/PostGIS, MySQL, MSSQL, Oracle, MongoDB and Firebase.
  • Processing GIS, remote sensing, meteorological and environmental datasets for operational use.
  • Lead delivery end to end: architecture through testing, deployment and production support.
Jan 2017 — Dec 2019Dhaka, Bangladesh

Application Developer

Navana Group

  • Built enterprise web applications with Python, Django, Angular, RxJS and TypeScript.
  • Developed reusable, responsive frontend components for large-scale business systems.
  • Built and integrated REST APIs and backend services in PHP/Laravel.
  • Designed and tuned database operations on Oracle and MSSQL.
  • Improved application performance by roughly 40% through code optimisation, lazy loading and better data handling.
  • Implemented authentication, business logic and third-party API integrations.
  • Handled debugging, testing, deployment support and production troubleshooting.

Systems in operation

Government and multilateral projects, delivered through RIMES with funding from the World Bank, UNDP, IFAD, USAID, NORAD and others. Several are still running in daily operational use.

Web platform

Flood Forecasting and Warning Decision Support SystemFFWC-DSS

Agency
Bangladesh Water Development Board, Ministry of Water Resources
Funding
World Bank, via the CARE project

A web platform for climate-smart flood management, supporting forecasting and decision-making across Bangladesh's water sector. It brings together real-time hydrometeorological data, transboundary water levels, Sentinel-1 flood extent maps and model forecasts in one place.

Trend analysis tools, interactive GIS mapping, automated bulletin generation and email alerting cut the time between a forecast arriving and a warning going out.

PythonDjango RESTAngularLeafletGeoPandasHighchartsPostGISNginxUbuntu
Web platform

Instant for South AsiaINSTANT-SA

Region
Afghanistan, Bangladesh, Bhutan, India, Maldives, Myanmar, Nepal, Pakistan, Sri Lanka
Funding
South Asia Hydromet Forum, facilitated by the World Bank

A temperature forecasting and climate visualisation platform for nine countries, built for heat monitoring and health-related early warning. Forecast data is pulled at six-hour intervals from the ECMWF HRES model and the UK Met Office system, and the backend computes daily maximum and minimum temperatures for mapping and time-series display.

Users switch between forecast source, initialisation, climatology, temperature indices and UTCI heat-stress layers, with percentile context showing how the coming days compare to the historical record.

DjangoReactdeck.glGeoPandasHighchartsPostGISNginxUbuntu
Web platform

Advanced Warning & Analytics for Risk & EmergenciesAWARE

Agency
Department of Disaster Management, Ministry of Disaster Management and Relief
Funding
IFAD, via the Provati3 project

A national multi-hazard decision support system covering floods, cyclones, tsunamis, landslides, earthquakes, lightning, flash floods and extreme heat, cold, wind and rain — used across preparedness, response, recovery and mitigation.

It carries a live alerts map graded by severity, emergency bulletins, inland riverport and marine warnings, 30-day incident tracking, accumulated rainfall by upazila, and a risk index with return periods. Hazard and service directories and emergency hotlines sit one tap away.

JavaScriptCodeIgniter 3LeafletGeoPandasHighchartsPostGISMySQLNginx
Web platform

Integrated Forecast Dissemination PortalINSTANT

Agency
Bangladesh Meteorological Department, Ministry of Defence
Funding
UNDP, ECHO, USAID, UCAR, Australian Aid

A forecast dissemination portal built under UNDP's Disaster Risk Management in Cox's Bazar project, with support from the Swiss Agency for Development and Cooperation, ECHO and the people of Japan, delivered with the Bangladesh Meteorological Department.

JavaScriptCodeIgniter 3LeafletGeoPandasHighchartsPostGISMySQLNginx
Web platform

National Livestock Advisory SystemNLAS

Agency
Department of Livestock Services, Ministry of Fisheries and Livestock
Funding
World Bank and USAID, via the CARE and SHOUHARDO-III projects

A livestock advisory and decision-support platform delivering climate-informed guidance to farmers and department officials. The framework and feature set were prototyped under the World Bank-funded CARE project with a technical working group drawn from BLRI, DLS, BAU and the SAARC Agriculture Centre.

AngularDjangoLeafletPostGISMySQLNginxUbuntu
Mobile app

আবহাওয়া (Abohawa)Official weather app of the Bangladesh Meteorological Department

Agency
Bangladesh Meteorological Department, Ministry of Defence
Funding
NORAD, via the BRIDGES project

The country's official weather app, carrying real-time forecasts, severe weather warnings, radar imagery and climate information for the whole of Bangladesh — the public-facing end of BMD's forecasting operation.

FlutterDartDjango RESTPostGISMySQLNginx
Mobile app

FFWC AppOfficial flood forecasting app of the Bangladesh Water Development Board

Agency
Bangladesh Water Development Board, Ministry of Water Resources
Funding
German Federal Foreign Office, via anticipatory action projects

Flood information and early warning in people's pockets: real-time forecasts for rivers and stations nationwide, live water-level charts, and location-based alerts that fire when risk rises nearby — at home, travelling or working in the field.

Interactive maps show impact zones and station detail, with short and long-range predictions on a single dashboard. A later release added an inundation map and an all-station view.

FlutterDartDjango RESTPostGISMySQLNginx
Mobile app

BAMISBangladesh Agro-Meteorological Information System

Agency
Department of Agricultural Extension, Ministry of Agriculture
Funding
World Bank, via the CARE project

Localised, science-based support for farmers adapting to a changing climate. It pairs real-time weather forecasts and flood alerts with crop advisories tailored to the user's location and AI-powered disease detection from a photograph.

FlutterDartDjango RESTYOLOv8TensorFlowPostGISNginx

Technical range

Frontend

React, Next.js, Angular, RxJS, TypeScript, JavaScript, HTML5, CSS3

Backend

Python (Django, Django REST, FastAPI, Flask), Node.js, Express, NestJS, PHP (Laravel, CodeIgniter)

Mobile

Flutter and Dart, native Android, Swift

GIS & remote sensing

PostGIS, ArcGIS, QGIS, GDAL, GeoPandas, Leaflet, Mapbox, deck.gl, spatial analysis, Sentinel-1 imagery

AI & computer vision

YOLOv8, TensorFlow, OpenCV, Scikit-learn, CNNs, LLM and LangChain integration, crop disease detection, landslide susceptibility modelling

Data

PostgreSQL, MySQL, MSSQL, Oracle, MongoDB, Redis, Firebase

Infrastructure

Docker, Kubernetes, Apache Kafka, RabbitMQ, Linux, Nginx, Apache, Jenkins, GitHub Actions, CI/CD, AWS, Azure, GCP, microservices

Education

  • B.Sc. in Computer Science & EngineeringDaffodil International University, 2012–2016 — GPA 3.35 / 4.00

Certifications

  • NWP Data Processing & VisualizationBangladesh Meteorological Department, 2020
  • Data Science & AI EssentialsCoursera, 2023

Languages

  • BengaliNative
  • EnglishProfessional working proficiency
  • HindiModerate