< Hello, I am />
Aayush Dangol
Computer engineering graduate specializing in Artificial Intelligence and Machine Learning. Building intelligent systems that solve real-world problems.
About Me
A research-driven software engineer motivated to deliver scalable, intelligent solutions.
I am a Computer Engineering graduate specializing in Artificial Intelligence and Machine Learning. My journey is rooted in building, training, and optimizing intelligent systems, with a strong focus on applied ML, LLM-based systems, and performance optimization.
I thrive on tackling complex architecture challenges—from optimizing end-to-end STS pipelines to designing graph-based knowledge models. My goal is to bridge the gap between AI research and secure, production-ready software.
Applied AI/ML
Hands-on experience in building, training, and optimizing intelligent systems with a focus on real-world scalability.
RAG Systems
Expertise in designing organizational memory systems using vector databases (Qdrant) and graph databases (Neo4j).
Cyber Security & AI
Combining intelligence with security, with recognition from the Dutch NCSC and CTF wins.
Experience
My professional journey in the world of AI and Software Engineering.
L1 Software Engineer (AI/ML)
Dec 2025 – PresentVivasoft Nepal
- Developed and enhanced agentic AI systems, improving orchestration and efficiency of LLM-based workflows.
Associate AI/ML Engineer
Aug 2025 – Dec 2025Vivasoft Nepal
- Worked on in-house AI systems, focusing on RAG, AI memory management, stress testing, and comparative evaluation of different AI models.
- Optimized end-to-end STS pipelines, improving efficiency and accuracy.
- Designed and implemented an AI-powered job interview agent.
- Deployed AI applications on GCP and AWS.
- Collaborated with overseas engineering teams on production-ready deployments.
AI/ML Trainee
Nov 2024 – Feb 2025Bajra Technologies
- 3-Month Traineeship focused on applied machine learning and model deployment.
Skills & Expertise
A comprehensive map of my technical proficiencies and specializations.
Programming
Frameworks & Tools
AI/ML Specialization
Featured Projects
Showcasing my work in AI, Machine Learning, and Cyber Security.
VivaDai — Organizational Memory Management System
Designed and developed an organizational memory platform to capture, structure, and retrieve institutional knowledge.
Key Details
- Built backend services with FastAPI and real-time collaboration using RTC.
- Integrated Gemini LLM for intelligent knowledge retrieval and contextual responses.
- Utilized Neo4j for graph-based knowledge modeling and Qdrant for semantic vector search.
Kothon Analytics Platform
Served as backend engineer for a scalable analytics platform.
Key Details
- Implemented microservices in Go and Python.
- Designed and maintained a microservice architecture for scalability and reliability.
- Managed deployment and operations on GCP, ensuring production stability.
- Optimized LLM token usage, reducing outages and improving inference efficiency.
Aakashwani & Capricorn — AI Outgoing Call Platform
AI-driven outgoing call platform with real-time voice interactions.
Key Details
- Developed backend services and admin dashboard using FastAPI and Jinja2.
- Integrated PJSIP with Gemini LLM via WebSockets to enable real-time AI-driven voice interactions.
- Delivered a comprehensive end-to-end calling platform, encompassing orchestration, AI processing, and administrative control.
Micrograd-Inspired Autodiff Engine
Developed an automatic differentiation engine for arithmetic operations.
Key Details
- Applied the engine to solve linear regression problems, demonstrating practical use of gradients.
CNN with Backpropagation from Scratch
Implemented forward pass for a multi-layer CNN and performed backpropagation manually.
Key Details
- Verified gradient correctness by comparing with PyTorch autograd.
BPE Tokenizer from Scratch
Built a tokenizer engine using the Byte Pair Encoding (BPE) algorithm.
Key Details
- Trained the tokenizer on an English text corpus (books), learning efficient subword representations.
- Evaluated for vocabulary coverage and tokenization efficiency.
Transformer for English–Nepali Translation
Implemented encoder-decoder architecture based on the original Transformer paper.
Key Details
- Trained the model on an English–Nepali parallel corpus for machine translation.
- Evaluated performance using BLEU scores and qualitative translation assessment.
Education
Bachelor’s in Computer Engineering
2020 – 2025Advance College of Engineering and Management
+2 Science
2018 – 2020Siddhartha Vanasthali Institute
Certifications & Awards
Recognition from the National Cyber Security Center (Dutch Government)
HexHimalayan CTF Winner 2023
2023
LOCUS Beginner CTF – Third Place (2023)
2023
Get In Touch
Have a project in mind or want to collaborate? My inbox is always open.
ayushazhar@gmail.com
Phone
+977-9810384095
Location
Kathmandu, Nepal