The Evolution from Static Dashboards to Intelligent Query Engines

Traditional business intelligence (BI) relies on rigid, pre-aggregated dashboards. When business executives need answers to non-standard questions—such as evaluating loan default correlations across localized dealership tiers—they must submit tickets to overloaded data engineering teams.

Modern data architecture bridges this gap by decoupling raw analytical computation from the query interface:

  1. Normalized Semantic Data Layer: Well-modeled relational schemas with comprehensive column metadata and referential integrity constraints.
  2. Context-Aware Natural Language Translation: Dynamic schema injection into language models to generate syntactically valid ANSI SQL queries.
  3. Strict Sandboxing & Security: Read-only transaction scopes, automated LIMIT clauses, and query cost timeouts to protect production database clusters.

Flagship Project: Prompt-Based Business Intelligence System

I engineered the Prompt-Based Business Intelligence System to eliminate analytical bottlenecks for non-technical leadership teams. The system converts plain-English queries into deterministic, optimized SQL statements, executes them against read-only replicas, and renders clean tabular charts within seconds.

System Deep-Dive

Prompt-Based BI Architecture

Complete engineering breakdown of the Text-to-SQL engine, automated schema pruning, read-only security sandboxing, and chart visualization rendering.

Read Architecture Deep-Dive →
Production Work

Selected Engineering Work

Comprehensive portfolio of data products, automation pipelines, and machine learning models shipped across production environments.

Explore Selected Work →

Academic Foundation in Data Analytics

I graduated Ranked 2nd with a 3.82 / 4.0 GPA in the Bachelor of Science in Data Analytics program at the Delhi Skill and Entrepreneurship University (DSEU). My curriculum combined rigorous mathematics with applied computing:

  • Statistical Foundations: Probability distributions, hypothesis testing (Z-tests, t-tests, ANOVA), parametric and non-parametric inference.
  • Multivariate Analysis: Principal Component Analysis (PCA), factor analysis, multi-collinearity diagnosis (VIF), and logistic regression.
  • Relational & Non-Relational Data Engineering: Advanced SQL (window functions, recursive CTEs, indexing strategies), Python (Pandas, NumPy, Scikit-learn), and data warehousing.