Projects Matrix

Scalable multi-tenant activity feed engine implementing Fan-Out-on-Write architecture for efficient user activity delivery.

Problem

Modern SaaS platforms require scalable activity feeds that efficiently deliver updates to many users while maintaining query performance.

Solution

Designed a multi-tenant activity feed engine using a Fan-Out-on-Write architecture, optimized MongoDB aggregation pipelines, and partial indexing to improve feed retrieval efficiency.

Result

Built a scalable database-driven backend architecture demonstrating NoSQL schema design, indexing strategies, and high-performance feed generation.

• Multi-tenant architecture
• Fan-Out-on-Write feed design
• MongoDB aggregation pipelines
• Partial indexing for optimized query performance

Browser-based packet sniffing dashboard for real-time protocol analysis and live port-to-service mapping.

Problem

Watching live network traffic usually means switching between CLI tools and manually mapping ports to services.

Solution

Built a Flask dashboard on top of Scapy that captures packets in real time, performs protocol analysis, and maps 20+ common ports to services, with a fallback simulation mode when capture permissions are insufficient.

Result

A single-pane browser view of live network activity that works both with and without elevated packet-capture privileges.

• 20+ ports mapped
• Real-time protocol analysis
• Graceful capture-permission fallback

Normalized Oracle PL/SQL inventory management system with automated business logic and relational database design.

Problem

Managing inventory efficiently requires maintaining data integrity while automating common business operations.

Solution

Designed a normalized relational database of 12 interconnected entities and implemented stored procedures, triggers, and views to automate business workflows while enforcing data consistency.

Result

Delivered a production-style relational database showcasing SQL, Oracle PL/SQL, normalization, and database automation techniques.

• 12 normalized database entities
• Stored procedures
• Database triggers
• SQL views

Server log parser that survives real-world log messiness and surfaces traffic patterns and anomalies.

Problem

Raw server logs are noisy, inconsistent, and full of malformed lines that break naive parsers.

Solution

Engineered a CLI + web dashboard log analyzer that gracefully handles 38 operational edge cases and tracks status codes, request counts, hourly traffic, anomaly flags, and malformed lines.

Result

A tool that turns messy access logs into a clean overview of usage patterns and anomalies without crashing on bad input.

• 38+ edge cases handled
• Status code + hourly traffic tracking
• Anomaly flagging

C++ code editor with a VS Code-inspired browser GUI, built from the data-structure level up.

Problem

Learning how editors actually work means building one — buffers, undo/redo, and recent-file management from scratch.

Solution

Implemented dynamic text-line storage via vector strings, a recent-file queue, and stack-based undo/redo history trackers, wrapped in a dark VS Code-inspired browser GUI.

Result

A working editor that demonstrates DSA choices (vectors, queues, stacks) applied to a realistic product surface.

• Vector-based line buffer
• Stack-based undo/redo
• Recent-file queue

Desktop recreation of the iconic endless runner, tuned for smooth collision detection.

Problem

Recreating an endless runner means keeping collision and asset loops fast enough to feel responsive at speed.

Solution

Built the game in C++, optimizing geometric bounding shapes and asset collision engine loops to keep the main loop tight.

Result

A playable desktop clone of the endless runner with smooth collision behavior and stable frame pacing.

Grid-based console Minesweeper with recursive reveal, difficulty scaling, and persistent high scores.

Problem

A faithful Minesweeper needs a correct recursive reveal, persistent scores, and clean difficulty scaling.

Solution

Built the game in C++ using structural arrays and file handling, adding a custom flagging system, high-score tracking files, dynamic difficulty scaling, and an automated recursive reveal algorithm. Developed with Ayesha Arif.

Result

A polished console Minesweeper that correctly cascades empty cells, persists scores across runs, and scales with difficulty.

• Recursive reveal
• Persistent high scores
• Dynamic difficulty