Technical & Executive Competencies

Expertise & Codebases

From corporate executive leadership at Yarsha Byte to foundational technical engineering. A unified view of my managerial and technical skill sets, alongside the projects that demonstrate them.

Managerial Skills

Corporate roadmap, team leadership, and strategy for Yarsha Byte

Core Competencies

Executive Leadership

Founding and driving corporate vision

Enterprise Architecture

Designing scalable multi-tier digital systems

Corporate Strategy

Client transformation & strategic growth

Team Management

Engineering mentorship & performance

Featured Systems & Repositories

Full-Stack & WebEnterprise production deployment

Avenue Butwal Platform

Premier enterprise digital ecosystem engineered at Yarsha Byte

Flagship digital platform designed and architected under Aashish's leadership at Yarsha Byte, featuring high-speed Next.js SSR architecture, fluid typography, and bespoke interactive components.

Architectural Highlights
  • Modular micro-component architecture with dynamic viewport rendering
  • Zero-overhead server-side rendering for optimal Core Web Vitals
  • Integrated brand design language and mobile touch-optimized layouts
Next.jsTypeScriptTailwind CSSEnterprise ArchitectureSEO Engine
View SourceLive System

Technical Skills

Full-stack web, AI, and systems engineering

Core Competencies

Next.js & React

Modern SSR, App Router & UI engineering

TypeScript & JavaScript

Strict type safety & web logic

Python & AI Engineering

Transformers, NumPy, & ML algorithms

C & Low-Level Systems

Direct memory management & algorithms

Featured Systems & Repositories

AI & LLMs100% Native NumPy, 0 ML Frameworks

Mini LLM From Scratch

Transformer-based chatbot built using pure NumPy and manual backprop

A tiny Transformer-based conversational model built from the ground up using exclusively NumPy. Implements multi-head self-attention, feed-forward layers, manual matrix backpropagation, and positional encodings without PyTorch or external deep learning frameworks.

Architectural Highlights
  • Custom multi-head scaled dot-product self-attention mechanism
  • Exact mathematical backpropagation for all linear and attention projection layers
  • Custom vocabulary tokenizer, cross-entropy loss, and temperature sampling inference loop
PythonNumPyTransformersManual BackpropMatrix Calculus
View Source
AI & LLMsSub-second semantic clinical retrieval

Clinical Doctor AI Assistant

Healthcare diagnosis and inquiry assistant powered by Hugging Face & ChromaDB

Medical intelligence assistant leveraging Hugging Face transformer models, ChromaDB vector database for semantic medical embeddings, and an intuitive reactive frontend interface for clinical query handling.

Architectural Highlights
  • Vectorized medical knowledge base indexed with ChromaDB
  • Semantic similarity search pipeline with Hugging Face embeddings
  • Rapid query disambiguation and patient context synthesis
PythonHugging FaceChromaDBVector EmbeddingsJavaScriptHTML5
View Source
AI & LLMsSpecialized intent classification

Pretrained Custom AI Agent

Autonomous conversational agent with fine-tuned domain intelligence

Modular AI agent architecture featuring domain-adapted pretrained weights, intent routing, and dynamic response generation for specialized automated interactions.

Architectural Highlights
  • Task-specific context adaptation and token constraint enforcement
  • Modular prompt templates and fall-back generation strategies
PythonAI AgentNLPModel Fine-tuningChatbot
View Source
AI & LLMsZero-latency intent parsing

Simple Chatbot NLP

Rule-based and pattern-matching conversational agent

Lightweight conversational agent exploring intent identification, pattern matching, token tokenization, and dynamic dialogue trees in Python.

PythonNLPPattern MatchingDialog Systems
View Source