Skip to content
View saminadamn's full-sized avatar
4 followers · 4 following

Block or report saminadamn

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
saminadamn/README.md

Samina Parveen

BIT Mesra CSE '27 · Software Engineer · Backend Engineer · Software Engineering Intern @Pulse

saminadamn Email LinkedIn

I build scalable backend applications, REST APIs, and data-driven software using Python, FastAPI, PostgreSQL, and modern cloud technologies. I'm interested in writing clean, scalable code and solving real-world engineering problems through reliable software. Currently exploring distributed systems, scalable APIs, and production-grade backend engineering.

Featured Projects

Projects

  1. Namma AI: Real-time, full-stack urban traffic intelligence platform for Bengaluru built for Flipkart Gridlock 2.0. Async FastAPI + PostgreSQL backend, a CatBoost model with <5ms inference for road-closure prediction, incident-aware OSRM routing (<3s route computation), and live WebSocket telemetry to a Next.js dashboard. Live demo
  2. Behavioral Attention Score (BAS) Framework: A LangGraph multi-agent research framework processing 10,000+ conversational data points to extract behavioral signals and classify attention states, with custom pipelines accelerating longitudinal tracking by 40%.
  3. DHANSETU — Financial Inclusion Platform: A secure full-stack platform (React, Node/Express, MongoDB, JWT) automating NBCFDC loan eligibility for 500+ simulated users, cutting manual processing time by 75%, powered by a KNN risk-banding ML pipeline. Smart India Hackathon 2025 Finalist.
  4. RepoGraph: Graph-based repository analyzer that reduces LLM token usage by 50–70% through dependency-aware code retrieval.
  5. Aurum AI: An autonomous enterprise cost intelligence system that analyzes operational data to detect cost inefficiencies, quantify financial impact, and recommend actionable optimizations. Built with Next.js, Tailwind CSS, and PapaParse for CSV-driven analysis. Live demo.

Achievements

  1. Smart India Hackathon 2025 — Grand Finalist (Dhansetu).
  2. Flipkart GRiD Gridlock 2.0 — built Namma Traffic with team ML Baddies.
  3. Amazon ML Summer School '26 — selected participant.

Open Source Contribution

HotSM (OpenStreetMap Toolkit)[https://github.com/hotosm/field-tm/issues] Contributed documentation and geospatial improvements to HotSM projects.

Merged / Active Contributions

  • docs: Excel & Power BI OData Visualization Guide
  • docs: QField Project Creation Workflow
  • fix(splitter): Handle Polygons with Connected Rings

Activity

Samina's contribution graph snake gif snake gif

Pinned Loading

  1. namma-traffic namma-traffic Public

    AI traffic intelligence platform for Bengaluru — predicts incident severity & road-closure probability with CatBoost, recommends officer/barricade deployment, auto-generates diversion routes, and s…

    TypeScript 2

  2. dhansetu2-live dhansetu2-live Public

    Loan-eligibility and credit-scoring platform for beneficiaries, officers, and channel partners. ML-assisted financial inclusion platform — SIH 2025 Grand Finals entry.

    JavaScript

  3. stratiq stratiq Public

    Enterprise BI & decision intelligence platform — multi-tenant auth/RBAC, dataset ETL, 4 analytics dashboards, ML-based churn/forecast/segmentation predictions, and a deterministic decision engine w…

    TypeScript

  4. ADHD-behavioral-attention-score ADHD-behavioral-attention-score Public

    A LangGraph-Based Multi-Agent Framework for Modeling ADHD Attentional Variability

    Python 1

  5. Repograph Repograph Public

    Graph-based GitHub repo analyzer that cuts LLM context costs 50–70% — builds a dependency graph via AST parsing, then selects only the most relevant files for a query instead of dumping the whole r…

    Python