ayush@engineering:~$ whoami

AI application & full-stack engineer
Focused on backend architecture, intelligent workflows,
document processing, and reliable software delivery.

ayush@engineering:~$ current_focus
> Omnitome
> AI-powered document intelligence
> Resource-aware processing & lifecycle reliability

ayush@engineering:~$ status
> Building | Testing | Learning | Shipping

Local-first AI reading & document intelligence

A local-first AI reading platform designed to transform large documents into focused reading experiences while preserving access to their original sources.

• Multiformat document ingestion and structured chapter extraction.
• Embeddings, reranking, retrieval, and AI-powered synthesis.
• Resource-aware processing and model lifecycle coordination.
• Persistent processing state, recovery workflows, and regression testing.

Core architecture

React + Vite → Node.js API → SQLite → Python FastAPI sidecar

Current focus: expanded document-corpus validation, reliability testing, and v1.0 release readiness.

Full-stack sales & pipeline management

A deployed CRM covering leads, clients, deals, tasks, sales activity, and pipeline analytics.

• React frontend and FastAPI backend.
• Database persistence and authenticated APIs.
• JWT authentication and role-based access control.
• Independently deployed frontend and backend.

Core architecture

React / Vite → FastAPI → SQLAlchemy → Database

Privacy-conscious browser AI

A Chrome extension exploring useful AI assistance without treating user privacy as an afterthought.

• Manifest V3 browser-extension architecture.
• On-device inference where supported, with controlled cloud-provider integration.
• Privacy modes and safeguards for webpage content and user interactions.

engineering_principles = {
    "architecture": "Understand the system before changing it",
    "reliability":  "Make failures observable and recoverable",
    "performance":  "Measure resource costs; optimize bottlenecks",
    "security":     "Respect identity, access, and data boundaries",
    "delivery":     "Validate behavior, not just successful builds",
}

I use AI-assisted development as an engineering tool, not a substitute for engineering judgment. I define problems, reason through architectural decisions, direct implementation, and review the resulting behavior.

• 🎓 Completing a BSc in Computer Science at the University of the People, with graduation expected in November 2026.
• 🧪 Developing practical machine-learning skills through coursework and hands-on exercises.
• 🧩 Deepening my understanding of backend systems, AI application architecture, testing, and performance.
• 🤖 Long-term interests: AI/ML systems and robotics.

I'm interested in teams building real products, solving meaningful technical problems, and giving engineers ownership of outcomes.