Skip to content
View rushammm's full-sized avatar
🎯
Focusing
🎯
Focusing

Block or report rushammm

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
rushammm/README.md

rusham elahi

ml · geo · cs grad · open to research roles & freelance

linkedin · email · portfolio · dblp

research

Healthcare access in south asia — ml study of healthcare infrastructure gaps across south asia, published at ICONIP 2024 (CORE A), co-first author. supervised by Dr. Usman Nazir (Oxford). · paper · arxiv

Deep Intelligent Urban Scoring — GNN approach to business site selection from POI graphs. represents urban regions as graphs (nodes = places, edges = geographic/functional proximity); the model learns embeddings that flag high-potential business zones, outperforming clustering and traditional spatial analysis on real city POI data.

Flood impact analysis — mapped 240,000 km² of sindh + south punjab flooding from sentinel-1 SAR in google earth engine; YOLOv5 on satellite tiles to detect schools/hospitals in flood zones; QGIS spatial intersection to quantify damage.

projects

Khabar — Pakistani news voice agent. you speak, it fetches Dawn headlines and reads them back over a live voice session. Next.js frontend + Python agent on the LiveKit Agents framework; Groq for STT (whisper-large-v3), LLM (llama-3.3-70b), and TTS in one provider. tool-calling for headlines / detail / category filter, with custom rate-limit + interruption rails to keep playback clean. · repo

Vision-metric — measures real-world width/height (mm) of a notebook from a single phone photo. iPhone 12 calibration (0.085 px reprojection error) → Mask R-CNN segmentation (val IoU 0.96) → homography on a credit card reference for mm scale. · repo

Healthcare Resource Predictor — predicts WHO Universal Health Coverage tier (SDG 3.8) from workforce + infrastructure data. Decision Tree, Random Forest, Gradient Boosting implemented from scratch in NumPy, validated 100% against sklearn. RF hits 78.46% on held-out 2020–2023. · repo

Locatepro — predicts business location suitability with XGBoost + satellite features (sentinel-1, landsat, google earth engine, NDWI, nighttime light). 12,000+ POIs from google places API. 80%+ accuracy, deployed. · live

Placify — internship management platform (t3 stack) connecting students, companies & universities.

tools i use

python typescript javascript sql next.js react node.js tailwind scikit-learn tensorflow pytorch xgboost pandas numpy google earth engine qgis postgresql git

open to ml engineering / research roles · published at ICONIP 2024 · based in Pakistan, open to remote

Pinned Loading

  1. ML-inDepth ML-inDepth Public

    ML algorithms implemented from scratch in Python, with small projects for better understanding.

    Jupyter Notebook 1

  2. Healthcare-Resource-Predictor Healthcare-Resource-Predictor Public

    Predicts a country's WHO Universal Health Coverage level (Low/Medium/High) from workforce and infrastructure data. Decision Tree, Random Forest, and Gradient Boosting built from scratch in NumPy

    Jupyter Notebook 1

  3. Vision-metric Vision-metric Public

    End-to-end CV pipeline that measures real-world width and height (mm) of a notebook from one phone photo using calibrated optics, Mask R-CNN, and a credit-card homography.

    Python

  4. khabar khabar Public

    Voice agent that reads Pakistani news from Dawn aloud. LiveKit Agents + Groq (Whisper, Llama 3.3, TTS) + Next.js.

    TypeScript