Live Application URL: https://bluestock-nifty-100-financial-intelligence-platform.streamlit.app/
The Bluestock Nifty 100 Financial Intelligence Platform (Project Nifty) is an institutional-grade equity research, automated financial analytics, and interactive intelligence portal engineered during a 45-day intensive development roadmap.
The platform processes, normalizes, validates, models, and visualizes over 11,000+ fundamental data points across Income Statements (P&L), Balance Sheets, Cash Flow Statements, Market Capitalization, and BSE filings for 92 Nifty 100 index constituent companies across a 10-year historical horizon (FY2015 – FY2024).
By integrating advanced financial ratio engines, machine learning clustering, natural language processing (NLP) pros/cons generators, valuation models, automated PDF tearsheet generation, and an interactive 8-screen dashboard, Project Nifty provides wealth managers, financial analysts, and investors with granular, data-driven equity insights.
The platform was built sequentially across 6 structured Sprints during the 45-day internship:
| Sprint | Days | Dates | Key Focus & Deliverables |
|---|---|---|---|
| Sprint 1 | Days 1–7 | June 13 – June 19 | Data Ingestion Engine, SQLite Database Schema, 16-Rule Data Quality Validator |
| Sprint 2 | Days 8–14 | June 20 – June 26 | 17 Ratio KPIs, 8-Pattern Capital Allocation Matrix, Composite Quality Rating |
| Sprint 3 | Days 15–21 | June 27 – July 3 | Dynamic Financial Screener, 11-Industry Peer Percentiles, Polar Radar Plots |
| Sprint 4 | Days 22–28 | July 4 – July 10 | 8-Screen Interactive Streamlit Web Portal, DCF & Relative Valuation Models |
| Sprint 5 | Days 29–35 | July 11 – July 17 | ReportLab PDF Publishing Engine (89 Company Tearsheets, Portfolio Summary) |
| Sprint 6 | Days 36–45 | July 18 – July 28 | Unsupervised K-Means ML Clustering, FastAPI REST API (16 Endpoints), 124 Tests |
- Ingested 7 core financial spreadsheets and 5 supporting datasets into a structured SQLite database (
data/nifty100.db). - Built a 16-rule Data Quality (DQ) Validator (
DQ-01toDQ-16) to catch negative sales, missing keys, date mismatches, and mathematical balance sheet discrepancies. - Standardized 92 constituent company tickers and normalized 10-year financial dates into uniform
YYYY-MMformats.
🔹 Sprint 2 (Days 8–14: June 20th – June 26th): Financial Ratio Engine & Capital Allocation Classifier
- Engineered a core KPI calculator computing 17 financial ratios (ROE, ROCE, OPM, NPM, D/E, ICR, FCF, Sales/PAT CAGR).
- Implemented the 8-Pattern Capital Allocation Matrix classifying companies based on Cash Flow signs (CFO, CFI, CFF) into strategic corporate archetypes (Reinvestors, Shareholder Returners, Debt-Funded Growth, etc.).
- Formulated a Winsorised Composite Quality Rating (
$0.30 \times \text{ROE} + 0.25 \times \text{FCF} + 0.25 \times \text{ROCE} + 0.20 \times \text{D/E}$ ). - Developed an Automated Column-Shift Auto-Healer that dynamically detected and healed 88 misaligned spreadsheet entries.
- Built a dynamic analyst screener with 15 customizable financial thresholds and 6 strategy presets (Quality Compounders, Value Picks, Growth Accelerators, Dividend Champions, Debt-Free Blue Chips, Turnaround Watch).
- Developed an 11-industry peer percentile ranking engine utilizing
PERCENTRANK.INCstatistical distribution. - Created 8-axis interactive Polar/Radar overlay charts comparing individual company scores against industry peer averages.
- Built and deployed the multi-page Streamlit web platform featuring responsive sidebar navigation, interactive Plotly charts, and custom styled CSS tiles.
- Embedded Discounted Cash Flow (DCF) intrinsic valuation and relative valuation models (P/E, P/B, EV/EBITDA target multiples).
- Implemented an NLP-based Qualitative Pros & Cons rule engine evaluating financial health signals into structured strengths and concerns.
- Built an automated PDF publishing pipeline combining ReportLab and Matplotlib graphics.
- Generated 89 individual company 2-page tearsheets, 11 sector intelligence summaries, and a Master Portfolio Report PDF.
🔹 Sprint 6 (Days 36–45: July 18th – July 28th): Machine Learning Clustering & FastAPI RESTful Backend
- Developed an unsupervised K-Means Clustering model categorizing companies into strategic financial clusters (validated via Elbow curves, Silhouette scores, and Pearson correlation heatmaps).
- Deployed a production-ready FastAPI backend (
src/api/main.py) exposing 16 REST endpoints with strict NaN-to-null JSON compliance. - Built a full automated unit and integration test suite featuring 124 passing test cases (
pytest).
The live web portal (bluestock-nifty-100-financial-intelligence-platform.streamlit.app) provides an 8-screen interactive workspace:
- Portfolio Macro Metrics: Highlights average portfolio ROE %, median P/E, median D/E, median 5-year Sales CAGR, total tracked constituents, and total debt-free blue chips.
- Sector Distribution: Interactive Plotly donut chart displaying the breakdown of Nifty 100 constituents across 11 broad sectors.
- Quality Score Leaders: Real-time leaderboards showcasing the top-ranked companies based on composite quality ratings.
- Search & Auto-Complete: Select any of the 92 constituent companies to inspect detailed financial profiles.
- 10-Year Trend Visualizations: Dual-axis financial charts comparing Revenue vs. Net Profit trends alongside ROE vs. ROCE trajectory.
- Qualitative Pros & Cons (NLP Engine): Rule-based natural language summary identifying fundamental strengths and risk warnings.
- 10 Granular Metric Sliders: Filter companies across ROE, D/E, FCF, Sales CAGR, PAT CAGR, OPM, P/E, P/B, ICR, and Quality Score.
- 6 Strategy Presets: One-click execution of analyst screeners (Quality Compounder, Value Pick, Growth Accelerator, Dividend Yield, Debt-Free, Turnaround).
- Export Capabilities: Clean tabular output with instantaneous CSV downloading.
- Industry Peer Matrix: Side-by-side metric comparison across 11 distinct sector peer groups with gold star benchmark highlights.
- Peer Percentile Rankings: Relative percentile ranks evaluating individual company positioning within its industry group.
- Polar Radar Chart Overlay: Interactive 8-axis radar plot mapping constituent strengths relative to peer group averages.
- Multi-Metric Comparison: Overlay multiple financial metrics (Sales, Profit, OPM %, ROE %, D/E, FCF) on a unified 10-year line graph.
- YoY % Growth Annotations: Real-time year-over-year growth percentage calculations and trend analysis.
- 3D Bubble Chart Visualization: Maps Revenue (X-axis) against ROE % (Y-axis) with bubble size representing Market Capitalization and colors denoting sub-sectors.
- Sub-Sector Summary Cards: Sub-sector median ROE and Quality Score group summaries.
- 8-Pattern Treemap: Interactive Plotly treemap grouping companies into 8 strategic allocation categories based on CFO, CFI, and CFF sign patterns:
- Reinvestor (
+ / - / -) - Shareholder Returns (
+ / - / -) - Liquidating Assets (
+ / + / -) - Distress Signal (
- / + / +) - Growth Funded by Debt (
+ / - / +) - Cash Accumulator (
+ / + / +) - Pre-Revenue / Burn (
- / - / +) - Mixed / Transitional
- Reinvestor (
- Pattern Filtering: Filter and inspect corporate constituents belonging to specific capital management profiles.
- BSE Filings Repository: Browse original BSE India Annual Report PDF links for all 92 companies across FY2015–FY2024.
- Live HTTP Verification: Validated links with browser headers to ensure immediate, error-free PDF downloads.
To evaluate company quality without distortion from extreme outliers, metrics are winsorised at the 10th (
The platform evaluates capital flow directionality across cash flow components:
Standardized feature vectors (
The platform enforces software reliability with 124 passing unit and integration test cases:
tests/api/test_companies_api.py .... [ 3%]
tests/api/test_health_api.py . [ 4%]
tests/api/test_screener_api.py .. [ 5%]
tests/api/test_sectors_api.py ... [ 8%]
tests/etl/test_loader.py .......... [ 16%]
tests/etl/test_normalise.py .......................................... [ 50%]
tests/etl/test_rules.py ......... [ 57%]
tests/kpi/test_cagr.py ....... [ 62%]
tests/kpi/test_cashflow.py ....... [ 68%]
tests/kpi/test_leverage.py ............ [ 78%]
tests/kpi/test_orchestration.py .... [ 81%]
tests/kpi/test_peer.py ... [ 83%]
tests/kpi/test_profitability.py ............... [ 95%]
tests/kpi/test_screener.py ..... [100%]
======================= 124 passed in 1.28s =======================
- Developer: Rishi Srivastava
- Project: Bluestock Internship — Project Nifty (45-Day Deliverable Challenge)
- Live App: https://bluestock-nifty-100-financial-intelligence-platform.streamlit.app/
- GitHub Repository: https://github.com/Mercer18/Bluestock-Project-Nifty