Project Details

Designed and developed an interactive HR Insights Dashboard using Power BI and Excel to analyse and visualise key human resources metrics. The dashboard provides executives and HR teams with a comprehensive view of workforce data, including headcount, retention, performance ratings, satisfaction scores, compensation trends, and training compliance. It features slicers for year and location filtering, dynamic visuals, and departmental breakdowns for granular analysis. Key insights include performance trends by tenure, the relationship between satisfaction and performance, and the identification of top and bottom performers. The solution supports strategic decision-making, improves visibility into employee engagement, and enhances operational efficiency across departments.

Live Interactive Dashboard

Project Details

This project demonstrates how explainable decision intelligence can be applied to healthcare waiting list management. Using synthetic but realistic patient referral data, the system engineers policy-aligned features such as delay ratios, breach flags, and severity bands to identify cases requiring operational review.

The solution prioritises transparency and governance over black-box prediction, with deterministic logic, fairness checks across age and region, and explicit ethical documentation. It is designed as a decision-support tool to assist clinical and operational teams, not to replace human judgement.

The project includes modular Python code, reproducible data pipelines, validation analysis, and visual outputs suitable for regulated HealthTech environments.

Project Details

This project presents a principal-level Trust & Safety and decision integrity framework for AI-assisted decision systems operating under adversarial conditions.

The platform models how decision systems can be manipulated through repeated interactions, threshold probing, automation boundary exploitation, and coordinated abuse. It demonstrates how behavioural abuse signals can be aggregated into interpretable risk scores, triggering proportional interventions such as progressive friction, automation downgrade, and specialist review.

Rather than focusing on model accuracy, the project emphasises system resilience, explainability, governance readiness, and safe degradation of automation. A controlled Python simulation illustrates how abuse can be detected early, harm contained, and human oversight preserved without penalising legitimate users.

The framework is designed as a platform-level reference for responsible AI, Trust & Safety operations, and decision governance.

Project Details

This project presents an AI Decision Governance & Oversight framework focused on ensuring transparency, accountability, and responsible use of automated decision systems.

The framework models how AI-assisted decisions can be logged, reviewed, overridden, and audited across their lifecycle. It introduces structured decision logs, risk-based governance rules, and human-in-the-loop escalation mechanisms to prevent unchecked automation and unmanaged harm.

Using a Python-based simulation, the project demonstrates how decision events, model metadata, inputs, and outcomes can be tracked consistently to support governance reviews, compliance needs, and operational oversight. Rather than optimising model performance, the emphasis is placed on explainability, traceability, and decision accountability.

This work is positioned as a platform-level reference for organisations deploying AI systems that require strong governance, regulatory readiness, and responsible decision-making controls.

Project Details

This project explores how blockchain principles can be applied to create a transparent, tamper-evident audit ledger for critical business decisions in regulated environments. Rather than focusing on cryptocurrency, the system uses cryptographic hashing and immutable event sequencing to record decision lifecycles such as creation, escalation, and resolution.

A hybrid architecture is demonstrated, where sensitive decision data remains off-chain while cryptographic proofs and audit metadata are recorded in an immutable ledger. This approach enables independent verification, traceability, and accountability without exposing confidential information.

The project includes decision classification, root-cause mapping, verification logic, and transparency metrics, supported by analytical dashboards and visual documentation. It is designed to support audit, compliance, and governance use cases rather than automate decision-making.

Project Details

This Excel-based project showcases a structured approach to analysing and visualising sales performance across countries, customers, and periods. Built for AtliQ Hardwares, the goal was to provide a clear, interactive snapshot of how different markets and client segments have grown or declined over the years, and how their performance compares to targets.

Key Highlights:

📌 Multi-Year Comparative Analysis: Track net sales from 2019–2021 by country and customer.

📉 Performance vs Target: Identify underperforming regions and visualise the gap between actual vs target sales using conditional formatting.

📈 Growth Metrics: Dynamic calculations of YoY growth percentages to spotlight top-gaining customers.

📊 Data-Driven Formatting: Heatmaps and colour-coded cells provide visual cues to support fast decision-making.

🔍 Filtering by Division & Market: Drop-down slicers make it easy to segment the data for focused insights.

Project Details

Business Insights 360 is a comprehensive Power BI dashboard designed to provide a multidimensional view of company performance across finance, sales, marketing, supply chain, and customer analytics. This project aimed to transform raw sales and operational data into clear, actionable insights for strategic decision-making.

Key Features:

Executive-Level Summaries: High-level KPIs including Net Sales, Gross Margin %, Net Profit %, and Forecast Accuracy across regions, customers, and product segments.
Interactive Slicers: Year, quarter, and market filters for on-the-fly exploration and analysis.
Performance Matrix & Trend Visuals: Visual breakdown of product and customer profitability, net error trends, and sales over time.
Segment-Level Drilldowns: Compare individual segment contributions and performance metrics like unit economics, gross margins, and risk profiles.
Forecast Accuracy Module: Track forecast vs. actual performance and identify key drivers of error.

Live Interactive Dashboard