Curriculum Vitae
Education
Kathmandu BernHardt College, Bafal, Kathmandu
Bachelor of Science in Computer Science and Information Technology | October 2019 – November 2024
Sainik Awasiya Mahavidyalaya, Phoolbari, Pokhara
Highschool (+2 Science) | 2010 – 2019
Technical Proficiencies
- Systems & Architecture: Linux (RHEL, Debian), KVM/libvirt, BIND9 DNS, ISC-Kea DHCP, LDAP/IdM, Kubernetes, OpenShift
- AI & Computer Vision: PyTorch, Mediapipe, OpenCV, Donut VDU, LangGraph, LLM Orchestration
- Cloud Ecosystem: Oracle Cloud (OCI Specialist), AWS, Azure, VMware ESXi, Prometheus, Grafana
- Software Engineering: Python, Java, C++, Bash Scripting, FastAPI, Selenium, Pytest, Terraform, CI/CD
Work Experience
QuickFox Consulting | Software Developer Intern – Automation & Computer Vision
Kathmandu, Nepal | Apr 2024 – Aug 2024
- Banking Automation: Contributed to the end-to-end Electronic Cheque Truncation and Clearing Pipeline for Global IME Bank Limited, automating cheque ingestion and processing in the Finacle Core Banking Solution using Selenium RPA bots.
- Visual Document Understanding (VDU): Integrated the transformer-based Donut model to extract structured cheque attributes, enabling Straight-Through Processing (STP) for high-confidence cases.
- Human-in-the-Loop: Designed a FastAPI-based verification interface that transitioned employees from manual entry to exception-based validation, improving auditability.
- Optimization: Refactored workflows with Pytest, reducing per-cheque ingestion latency by 70% (from ~40s to 10–12s).
- Synthetic Data Generation: Developed DocSim for Nepali, a template-driven framework using Pillow and OpenCV to generate high-fidelity synthetic cheques and forms to address data privacy constraints.
- Signature Verification: Prototyped an offline handwritten signature verification model using an Improved Inverse Discriminative Network.
Systems & AI Projects
Embedded Linux System with Daemonized TCP Service | Jan 2026
C, Buildroot, POSIX, QEMU
- Built a minimal AArch64 Linux image using Buildroot with reproducible build configuration and cross-compilation.
- Implemented a daemonized TCP socket service in C using POSIX APIs, supporting newline-delimited requests and persistent storage.
- Integrated service lifecycle management, including syslog logging, clean shutdown on SIGINT/SIGTERM, and safe file descriptor handling.
- Wired the service into system startup using init scripts and validated reliability across reboots, restarts, and failure scenarios using QEMU emulation.
Distributed Database & Observability Stack | Dec 2025
MongoDB, Prometheus, Grafana, KVM
- Deployed a 3-node MongoDB Replica Set on Debian 13 with a dedicated /25 subnet.
- Implemented secure intra-cluster communication using Symmetric Keyfile Authentication and RBAC.
- Developed an observability pipeline integrating Prometheus and Grafana with Percona MongoDB Exporter.
- Automated log rotation and retention using the Linux
logrotateservice.
Enterprise Linux Infrastructure & Identity Stack | Nov 2025
ISC Kea, Kerberos, LDAP, Red Hat IdM
- Architected a hybrid-distro lab on KVM/libvirt simulating an enterprise environment.
- Engineered an IPAM solution by integrating ISC-Kea DHCP with BIND9 for automated DDNS.
Agentic PR Orchestrator & MCP Server | Jun 2025
Python, LangGraph, Ollama, Slack API
- Built a state-machine LLM agent using LangGraph to orchestrate code analysis, staging, and PR creation.
- Engineered a custom Model Context Protocol (MCP) Server for secure local file system and shell execution.
- Created a FastAPI webhook server with Cloudflared Tunnels for real-time GitHub event feedback.
Multimodal Agentic System for GAIA Benchmarking | May 2025
LangGraph, Qwen3, GPT-4o
- Built an API-driven autonomous agent that executes multi-step reasoning for remote benchmark tasks.
- Designed a tool-augmented pipeline for web search, document parsing (PDF/CSV/Excel), and vision analysis.
- Achieved 85% accuracy on Level 1 tasks on the GAIA benchmark.
American Sign Language Recognition System | December 2023 – March 2024
Python, PyTorch, PyQt6, Pyttsx3, Mediapipe
- Developed a CV application with PyQt6 GUI.
- Engineered all CNN layers from scratch for gesture recognition.
- Used Mediapipe for hand coordinate tracking; created a custom dataset with skeleton images.
- Integrated Pyttsx3 for audio output of model predictions.
Face Recognition and Attendance System | May 2023 – June 2023
Python, OpenCV, Firebase, cvzone
- Built a face recognition and attendance system using dlib and Firebase Realtime Database.
- Implemented facial data encoding, live video capture, and dynamic attendance updates.
- Developed a user-friendly interface with cvzone.
Inkwell: Online Bookstore | April 2023 – September 2023
SQL, PHP, XAMPP, JS, Bootstrap, jQuery, HTML5, MariaDB
- Designed and developed an e-commerce platform for books.
- Implemented user profiles, purchase history, and book descriptions.
- Developed admin features for inventory management with MariaDB backend.
Specialization Coursework
-
Building Linux Servers Specialization Pearson Education Verify -
Cloud Computing: Infrastructure & Platforms Illinois Institute of Technology Verify -
Google Cloud AI Infrastructure Google Cloud Training Verify -
DevOps, DataOps, and MLOps Duke University Verify -
Cloud Native Development with OpenShift and Kubernetes Red Hat Verify
Professional Certifications
-
MongoDB Certified Associate Atlas Administrator MongoDB Verify -
OCI 2025 Certified Networking Professional Oracle Cloud Verify -
OCI 2025 Certified Observability Professional Oracle Cloud Verify -
OCI 2025 Certified Architect Associate Oracle Cloud Verify
Technical Skills
- Languages: Python, C/C++, C#, Java, SQL, JavaScript, PHP, HTML/CSS
- Frameworks: Flask, Django, ASP.NET
- Developer Tools: Git, Firebase, VS Code, Visual Studio, PyCharm, IntelliJ, pip3
- Libraries: PyTorch, TensorFlow, Pandas, NumPy, Matplotlib, Keras, Mediapipe, PyQt6, Tkinter
- Databases: MySQL, PostgreSQL, Redis
- Container Platforms: Docker, Kubernetes, Red Hat OpenShift
- Cloud Computing: Azure AI, IBM Cloud, Google Cloud Platform