Pramish Thapa

I am a

About

Back-end engineer building reliable, cloud-native services.

Back-End Software Engineer

2.5 years of production back-end experience in Java and Spring Boot on Google Cloud Platform, including credit risk and identity verification systems for Equifax (via Oncore Software Solutions).

Role: Back-End Software Engineer
Specialism: Java, Spring Boot, GCP
Location: United Kingdom

Degree: MSc Software Engineering (Distinction)
University: University of West London
Availability: Open to opportunities

I have delivered credit risk and identity verification systems for Equifax, and hold an MSc in Software Engineering with Distinction from the University of West London. I enjoy designing clean APIs, writing well-tested code, and shipping to the cloud.

Tech Stack

Languages, frameworks, libraries and platforms I have studied and deployed.

Languages

Java

Python

JavaScript

PHP

HTML5

CSS3

Front End

React

Angular

Vite

Tailwind CSS

Bootstrap

jQuery

Back End

Spring Boot

Node.js

Express

Flask

Databases & Messaging

MySQL

PostgreSQL

MongoDB

Firebase

SQLite

Apache Kafka

Cloud & DevOps

GCP

AWS

Docker

Kubernetes

Jenkins

Vault

Heroku

Vercel

Machine Learning

scikit-learn

PyTorch

pandas

NumPy

Tools

Git

Jira

Curriculum Vitae

A summary of my experience and education.

Pramish Thapa

Back-End Software Engineer · Java · Spring Boot · Google Cloud Platform

Back-End Software Engineer with 2.5 years building production microservices in Java and Spring Boot on Google Cloud Platform, including 1.5 years on Equifax's credit-risk and identity-verification platform. MSc Software Engineering with Distinction from the University of West London.

Download CV (PDF)

Professional Experience

Software Engineer

Oncore Software Solutions · client: Equifax Sept 2021 - June 2023

Java · Spring Boot · GCP Cloud Run · Pub/Sub · Firestore

Built and maintained Java/Spring Boot microservices on GCP Cloud Run for Equifax's credit-risk and identity-verification platform, working with US-based senior engineers across timezones.
Designed Cloud Pub/Sub publishers and subscribers for asynchronous service-to-service messaging, persisted to Firestore, and integrated third-party verification APIs with retry and error handling.
Deployed through Jenkins CI/CD, managed secrets with HashiCorp Vault, and configured IAM roles with SRE teams; fixed a production British-Summer-Time date-filtering bug.
Wrote JUnit/Mockito tests to 95%+ coverage and enforced quality/security gates with SonarQube and Fortify.

Software Engineer

InfoxIT Pvt. Ltd. Jan 2020 - June 2021

Java · Spring · Hibernate · MySQL

Developed back-end modules for an ERP platform used by banks and educational institutions - payroll automation, invoice generation, and inventory tracking.
Designed MySQL schemas and ORM mappings and optimised SQL queries for improved performance.
Worked across the full SDLC in Agile Scrum: requirements, sprint planning, code reviews, and client demos.

Education

MSc Software Engineering

Distinction 2023 - 2024

University of West London

Dissertation: Digital Twin & IoT for Smart Building Energy - ML pipeline (Random Forest, Linear Regression) on real-time IoT data.

BSc Computer Information Systems

2:1 2017 - 2021

Pokhara University, Nepal

Certifications

Full-Stack Development Programme - IT Career Switch
DevOps Engineering Programme - ThinkCloudly
Full-Stack Engineer Career Path - Codecademy
Neural Networks and Deep Learning - DeepLearning.AI

Languages

English (fluent) · Nepali (native) · Hindi (fluent) · Spanish (basic)

Research Experience

Battery Remaining Useful Life & Degradation Modelling

University of West London 2025 - present

Independent research supervised by Dr Fateme Dinmohammadi · Python · scikit-learn · PyTorch

Benchmarked five degradation models (linear, polynomial, exponential decay, random forest, Gaussian Process Regression) on the Oxford and NASA PCoE datasets, using chronological splits to avoid data leakage.
Built and deployed a live diagnostic dashboard for state-of-health tracking, stress analysis, degradation and end-of-life estimation.
Extending the work to temporal deep learning (CNN-LSTM, Transformer); cross-dataset benchmark paper in preparation for Sensors (MDPI).

Deep Learning for Turbofan Remaining Useful Life (NASA C-MAPSS)

University of West London 2025 - 2026

Temporal prognostics · CNN-LSTM · SHAP

Trained a CNN-LSTM hybrid on the NASA C-MAPSS FD001 dataset achieving RMSE 14.84 on the held-out test set with no overfitting, and applied SHAP for per-sensor interpretability.

Digital Twin & IoT for Smart Building Energy (MSc Dissertation)

Distinction 2024

Supervised by Dr Fateme Dinmohammadi · Python · scikit-learn · Flask · React

Built an end-to-end machine-learning pipeline predicting building energy consumption from real-time IoT data, combining Digital Twin technology with Random Forest and Linear Regression models.
Benchmarked four models across MAE, MSE, RMSE and R²; the best (Small Random Forest) reached R² = 0.9989, served via a Flask REST API and a React dashboard deployed on Heroku.