Vinayak Nautiyal

Problem

Building a secure, structured system for managing products, users, and transactions.

Key Work

• • Designed full-stack architecture using Next.js and PostgreSQL
• • Implemented authentication, RBAC, OTP verification, and rate limiting
• • Built product catalog, cart, and order workflows
• • Structured backend with Prisma and clean API patterns

Outcome

Delivered a production-style system with secure flows and scalable architecture.

Problem

Efficiently querying and analyzing large datasets.

Key Work

• • Built query engine for large CSV datasets
• • Implemented filtering, aggregation, and indexing logic
• • Designed data ingestion and processing pipeline
• • Optimized memory usage for large-scale data handling

Outcome

Enabled efficient analysis of large datasets with structured processing pipelines.

Key Work

• • Used MobileNetV2 for feature extraction
• • Applied PCA for dimensionality reduction
• • Trained SVM classifier on embeddings
• • Built preprocessing and evaluation pipeline

Outcome

Achieved ~92% recognition accuracy with optimized computation.

Problem

Google Pay exports are raw HTML with no structured categorization or usable analytics format.

Key Work

• • Built HTML parsing pipeline using BeautifulSoup and regex extraction
• • Implemented data cleaning and feature engineering using Pandas
• • Designed merchant enrichment using Google Places API with fuzzy matching
• • Added caching layer to reduce repeated API calls and improve performance

Outcome

Converted unstructured transaction logs into structured datasets with categorized spending and location context.

Problem

Existing trackers lack unified data tracking and meaningful feedback across habits, goals, and wellness metrics.

Key Work

• • Built full-stack Flask application with SQLite-backed data models
• • Implemented EMA-based habit tracking for consistency scoring
• • Developed server-side chart generation using Matplotlib (Base64 rendering)
• • Integrated AI-based routine analysis with fallback logic for reliability

Outcome

Delivered a self-contained analytics dashboard combining quantitative tracking with AI-assisted feedback.