Ashay Kushwaha — Systems Builder
Intend.Design.Deliver.
Ashay Kushwaha — Systems Builder

I build systems
that make complexity usable.

나는 센티넬 사이퍼다.
A CURIOUS MIND
WITH THE DISCIPLINE
TO TURN IDEAS
INTO REALITY.
Project Links
NSE Sentiment — live price + news NLP KarmaMap — NGO-volunteer matching platform NSE Risk Scanner — VaR / Monte Carlo
See My Work About Me
About

I turn messy data into tools people run.

I engineer open-source systems that turn raw data into decisions — NGO-volunteer matching platforms, climate policy dashboards, causal inference engines, real-time sentiment pipelines. Each project starts with a question and ends with a tool anyone can use.

29
Open Source Projects
Projects
68
PR Contributions
Contributions
50
PRs Merged
Merged
Why I Build

I believe the best tools should belong to everyone. Not locked behind paywalls. Not hidden in proprietary systems. Open-source is how we democratize the best tools — and that's exactly what I build.

My work spans cybersecurity risk analysis, climate policy visualization, causal inference, and social impact platforms. The common thread: taking complex data and making it actionable for the people who need it most.

What I bring
Full-stack data engineering — Python to ReactML & causal inference for real-world problemsOpen-source as an ethos, not an afterthoughtClean architecture, reproducible results
Interests
Cybersecurity AnalyticsClimate TechCausal InferenceDeveloper ToolingOpen-Source EcosystemsSocial Impact
Full-Stack EngineeringData & AIOpen SourceApplied Impact
Ashay Kushwaha
Stack

What I build with.

The instruments behind the systems — no percentages, just proof.

Languages
Python
Data, ML, automation
TypeScript
Frontend & tooling
SQL
Analytics & PostGIS
Frameworks
React
Interactive dashboards & PWAs
FastAPI
High-performance APIs
Streamlit
Rapid data apps
Node.js
Backend services
Data & AI
XGBoost
Risk scoring & prediction
Gemini
Content analysis & generation
VADER
Sentiment analysis
Causal Inference
Time series & treatment effects
Infrastructure
PostGIS
Spatial queries & mapping
PWA
Offline-first web apps
REST APIs
API design & integration
Docker
Containerized deployments
Selected Work

Shipped, tested, open-source.

Open-source, MIT-licensed. Each project ships as a complete, production-ready tool — not a demo.

Featured — Streamlit · Python

NSE Sentiment Analyzer

Live NSE price + multi-source sentiment

Enter any NSE ticker for a BULLISH / NEUTRAL / BEARISH signal: smart ticker search (504 aliases, handles rebrands and splits), 9-source news sentiment with event-aware scoring across 19 event types, a SmartScore 0-100 fusing recency-weighted EWMA, headline breadth and news volume, enhanced VADER plus a 123-term Indian financial lexicon, and RSI(14)/MACD technicals. 139 tests, AGPL v3.

StreamlitPythonNLPFinance
View on GitHub
KarmaMapHyper-local NGO-volunteer matching platformProduction-grade PWA connecting NGOs with nearby volunteers. PostGIS proximity matching scored by skill and road distance (OSRM), role-based dashboards for volunteers/NGOs/corporates, a karma points + streaks incentive system with leaderboard, offline-ready installable PWA, and Supabase Row-Level Security. React + Vite frontend on Vercel, Express backend on Render.TypeScriptNSE Portfolio Risk ScannerInstitutional-grade risk analytics for NSEUpload a holdings CSV (or try the sample) for a full risk report: Historical, Parametric and Cornish-Fisher VaR at 95/99, CVaR, Sharpe and Sortino, beta vs Nifty, 10,000-path Monte Carlo, HMM regime detection, Hierarchical Risk Parity optimization and scenario stress tests. 355 tests, 90 percent coverage, zero API keys. Deployed on Streamlit Cloud.Python355 testsFII/DII DashboardInstitutional flow tracking from NSE IndiaAuto-fetches daily FII/DII flows from NSE India into SQLite; 4 interactive charts (net flow trend, FII vs DII, rolling averages, Nifty overlay), date-range filtering and CSV export. Premium UI with Lucide icons, zero cron, no API keys. 49 tests, AGPL v3.Python49 testsAGPL-3.0DeltaGridParis Agreement NDC progress trackerComputes the gap between Paris Agreement NDC pledges and actual energy-transition trajectories for 200+ countries. A 0-100 Green Score from 6 energy shares, real-time slider re-ranking, Plotly choropleth maps, CSV/XLSX upload with auto-preprocessing, and country classification (hidden champions to laggards). 123 tests, MIT.Python123 testsMITCausalLensCausal inference for time seriesDid that policy work? Five causal methods — ARIMA ITS, SARIMAX, Bayesian STS (Google CausalImpact), Difference-in-Differences, and Synthetic Control — with counterfactual charts, p-values, 95 percent CIs, placebo sensitivity checks, and PDF/HTML export. 207 tests, 11 pre-loaded datasets, MIT.Python207 testsMIT
Activity

Contribution graph.

Contact

Got a problem worth solving?

Open to collaborations, research partnerships, and conversations at the intersection of data, security, and social impact.

Find me online.

All my work is open-source and MIT-licensed. I'm most active on GitHub and LinkedIn.

GitHubXLinkedInMedium
Let's talk.

Interesting project, research collab, or just want to geek out about causal inference? I'm here for it.

Connect on LinkedIn