Final-year Computing Science student at the University of Glasgow.
I build backend services in Python and the cloud infrastructure they run on:
APIs, caching layers, auth and data pipelines on Google Cloud.
// simulated traffic through my Redis write-through cache
GraduatingJuly 2027 · BSc (Hons)
Right to workUK · EU citizen
LanguagesPython · Java · SQL
CloudGCP · BigQuery · Redis
~/glasgow $ whoami
01 · About
Hi, I'm Arvid.
I'm a Swedish Computing Science student in Glasgow who likes building software that runs in the cloud.
Most of what I know I learned by deploying things. I've written Flask services on App Engine,
put Redis in front of MySQL as a write-through cache, built a JWT auth service
my other apps share, and spent a summer building a log pipeline into BigQuery.
Python is my main language and I also write Java. Security is a hobby of mine, which is why my projects
tend to include auth and logging.
Looking for
Graduate software engineering and cloud engineering roles where I'm stretched and keep learning.
02 · Experience
Building a log pipeline on GCP.
Infrastructure & Security Summer Analyst
Advanced Voice Research Labs, Inc. · Internship
Jun – Aug 2026
Project MELT · the log pipeline I helped buildGCP · PoC
01 · CollectFlask appsDrop-in logging module
02 · ShipCloud LoggingFrom GCP PaaS
03 · QueryBigQuerySQL for behaviour patterns
04 · SeeData StudioSpotting unusual users
Project MELT is AVRL's new platform for centralising application and request-level logs.
The data feeds their firewall rules and in-house monitoring.
What I did
Contributed to the drop-in Flask logging module
Built the pipeline proof of concept into BigQuery
Wrote SQL on BigQuery to find behaviour patterns
Used Data Studio to surface unusual traffic
What the data showed
Web app scannersautomated probing
Users running botscustomers automating access
Offshore trafficunexpected regions
Offered letters of recommendation by AVRL's Director of Engineering and Senior Director of Operations.
03 · Projects
Live systems, not screenshots.
Both are deployed on Google Cloud and share one login, so you can try them yourself.
Reads and writes go to a MySQL database with Redis as a write-through cache. One user can read and write,
another can only read. Those scopes come from logging in through auth.arvid.engineering.
Redis
MySQL
HAProxy
VPC
Compute Engine
What I learned
Redis on GCP Memorystore, reachable only inside a VPC
Running MySQL on an Ubuntu VM
Setting up HAProxy as a reverse proxy on Debian, plus VPCs and firewall rules on GCP (with AI assistance)
How it worksWrite-through
ClientLogged-in userScope from auth.arvid.engineering
Debian VMHAProxy reverse proxyPublic entry point
AppWrite → cache + DB togetherread: Redis hit, else MySQL
Memorystore · VPCRedisFast reads
Ubuntu VMMySQLSource of truth
Animation built with Claude Opus 5.5 from my repo, prompted by me.
An auth service that issues signed JWTs carrying access scopes. It publishes a JWKS endpoint so any service can
verify a token, and sets a cross-subdomain cookie so the other apps on arvid.engineering can see the session.
Python
Flask
JWT
JWKS
App Engine
What I learned
Running Python on App Engine (GCP PaaS)
Generating key pairs to sign JWTs (AI helped with the utility functions and UI)
Session cookies with Flask, scoped across subdomains
BrowserCookie on .arvid.engineeringVisible to every subdomain
Resource serviceVerify signatureFetches the public key
PublicJWKS endpoint/.well-known/jwks.json
Animation built with Claude Opus 5.5 from my repo, prompted by me.
04 · Toolkit
What I work with.
Languages
Python
Java
SQL
Backend
Flask
REST APIs
Sessions & cookies
JWT auth
Git
Cloud & infra
App Engine
Compute Engine
Memorystore
VPC
HAProxy
Linux
Data
MySQL
Redis
BigQuery
Cloud Logging
Data Studio
On the side: I'm a keen amateur in security. I've played with OAuth 2.0, JWT/JWKS and firewall rules, and I like knowing how the systems I build could be attacked.
Available for 2027 graduate roles
Let's build something reliable.
If you're hiring graduates for software or cloud engineering, I'd like to hear from you. Email is the fastest way to reach me.