Computer Engineer

about

Profile

I build and study distributed systems, networking, and cloud infrastructure, with a strong interest in how large-scale services are designed, deployed, secured, and operated in production environments. My focus lies at the intersection of backend engineering, infrastructure automation, and network security, where reliability, performance, and scalability are essential. I enjoy solving complex systems problems and continuously deepening my understanding of the technologies that power modern infrastructure.

technologies

Go
Next.js
Docker
Kubernetes
OpenTelemetry IconOpenTelemetry
Grpc Streamline Icon: https://streamlinehq.comgRPC

projects

Fraud Detection Pipeline

Fraud Detection Pipeline

Built a real-time fraud detection pipeline using Kafka and Spark Structured Streaming to run LightGBM inference and flag fraudulent activity.

SparkKafkaLightGBMCassandraTrino
View on GitHub →
Watson

Watson

Built a scalable security monitoring platform with Go microservices, PyTorch-based UEBA anomaly detection, and mTLS-secured communication using PKI.

GoPyTorchInfluxDBgRPCmTLS
View on GitHub →
Scheduler

Scheduler

Built a MERN stack platform for job listings and interview scheduling with real-time chat and secure JWT authentication.

MongoDBExpressReactNode.jsJWT
View on GitHub →
GestureBasedController

GestureBasedController

React Native app enabling PC control via smartphone IMU gestures, using sensor fusion and WebSocket for real-time commands.

React NativeIMU SensorsSensor FusionWebSocket
View on GitHub →

research

Agent-Based End-Host Monitoring in IPv6 SDN Intranets: Hybrid ML with UEBA

Oct. 2024 – May 2025

Proposed a scalable, low-overhead agent-based monitoring system for SDN-enabled IPv6 intranets. Deployed lightweight agents to collect time-series system and process metrics. Used mTLS-protected gRPC streams for secure transmission. Hybrid UEBA model with Isolation Forest + LSTM Autoencoders for anomaly detection. Integrated with SDN controller for dynamic flow rule enforcement; achieved sub-second telemetry latency.

UEBALSTM AutoencodersIsolation ForestSDNgRPCIPv6

CDD O-RAN: Causal Discovery-Driven MBRL for xApp Conflict Mitigation in O-RAN

Ongoing Research

Proposed CDD O-RAN for automated xApp conflict management in near-RT RIC. Learned offline causal NCP-KPI graphs using EMA-smoothed conditional mutual information. Reused the graph for conflict detection and causal model-based planning. Eliminated predefined QoS mappings and improved Out-Of-Distribution planning performance over dense MLP world models.

O-RANxAppsCausal DiscoveryModel-Based Planning

experience

Technical Contributor

2024 – 2025
Engineers Without Borders

Lalitpur, Bagmati

  • Contributed to club activities by supporting technical projects and initiatives
  • Automated event coordination and streamlined communication workflows to improve operational efficiency.

Program Assistant

Mar. 2023 – Apr. 2025
IEEE, Pulchowk Branch

Lalitpur, Bagmati

  • Assisted in the organization of IEEE Xtreme 17.0 Global Competitive Competition, managing the Nepal segment

Software Developer

Jul. 2022 – Jun. 2024
PDSC

Lalitpur, Bagmati

  • Led the successful conversion of a traditional website to a fully responsive React-based platform
  • Contributed to the development of a MERN stack-based streaming platform
  • Managed a workshop program centered on blockchain technology, overseeing sessions and logistics