I build and scale the backends behind products people rely on, from the first commit to tens of thousands of users.
I’m a VP at Bank of America, building Python trading tools for the XVA desk. Before that I spent six years at Bakken & Bæck in Oslo, taking startup products from zero to production.
Senior Front Office Developer (VP), XVA Technology, Bank of America
Based in
London, UK
Email
hello@jeraiz.co.uk
10+yearsbuilding and scaling backend systems
3productsbuilt from the ground up and taken to production
140,000+peoplenow use Kron, the investment platform whose first backend I built
DOGAAwardfor Samsolgt, from Design and Architecture Norway, 2020
01
Selected work
Three products from my six years at Bakken & Bæck, and one research project.
Case study · 1 of 3
Kron
An investment platform for everyone
140,000+people use Kron today
Source: kron.no, October 2026
Role
Backend lead
Years
2016 – 2019
For
Formue
What it is
A Norwegian investment platform that makes saving in funds simple: a personalised portfolio matched to your risk tolerance and time horizon, with automated advice and low fees. It was launched in 2017 by the wealth manager now called Formue.
Fig. 1 Web and mobile clients and the Preact admin dashboard called the Tornado API, which kept its data in PostgreSQL. The Pandas portfolio analytics ran inside the service, and Formue executed the trades.
Request path
Third-party system
What I did
Built the backend from scratch, on my own: a Python (Tornado) service on PostgreSQL.
Implemented the API integrations with Formue, Kron’s parent company at the time, which executed the trades.
Built the admin dashboard in Preact.
Used Pandas for statistical analysis of each client’s portfolio.
8AWS environments managed as infrastructure-as-code
Role
Backend engineer
Years
2019 – 2022
For
DNB Eiendom
What it was
DNB Eiendom’s digital, fixed-price way to sell a home, launched in 2019. Sellers did the preparation and held the viewings themselves, an estate agent handled the bidding and the legal side, and every task, document and contract lived in one place online.
Fig. 2 Requests from the React app entered a VPC at a load balancer and passed through the Nginx ingress controller to the Python REST APIs, which ran on Kubernetes on EC2 and kept their data in MySQL on RDS, also inside the VPC. S3, Route 53 and Secrets Manager sat alongside. CloudFormation defined the eight environments (two development, four test, two production), Jenkins built, tested and deployed to the cluster, and privileged access into the VPC went through a remote-access gateway and a credential vault.
Request path
Deployment and access
What I did
Designed and implemented the Python REST APIs that kept the state of every home sale in sync across the app.
Liaised with DNB, Norway’s largest financial services group, to architect a cloud setup with bank-grade security.
Ran eight AWS environments (two development, four test, two production) as infrastructure-as-code.
Set up continuous integration with Jenkins, and used Kubernetes for performance, scalability and availability.
Worked in a team of engineers and designers from Bakken & Bæck and DNB.
Recognition
DOGA Award for Design and Architecture, 2020
Design and Architecture Norway (DOGA) · Interaction design
“A good piece of service design that makes the whole sales process more transparent.”
From the jury’s statement, translated from Norwegian
Awarded to DNB Eiendom and Bakken & Bæck. DOGA calls the award Norway’s most coveted distinction for design and architecture projects. I’m credited on the project team as backend developer.
Samsolgt ran until 2024. DNB told E24 in June 2024 that much of the technology built for it lives on across DNB Eiendom, which handles around 20,000 home sales a year.
A Norwegian circular refill service, set up by Orkla, Æra and Bakken & Bæck to cut plastic waste. Soaps, detergents and cleaning products arrived in reusable containers; empties were left outside, collected, cleaned, refilled and returned.
Fig. 3 Requests from the React app passed through a load balancer and the Nginx ingress controller to the Tornado API, which ran on Kubernetes on Linode, kept its data in PostgreSQL and exchanged pickup and delivery orders with Helthjem’s logistics API. Jenkins ran the tests and deployed to the cluster.
Request path
Deployment
Third-party system
What I did
Built the backend from scratch, on my own: a Python (Tornado) service on PostgreSQL.
Integrated Helthjem’s logistics API so that pickups and deliveries of containers were automated.
Set up CI/CD with Jenkins to run the tests and deploy to Kubernetes.
Hardened the Nginx layer with the right security headers.
I researched how WaveNet, DeepMind’s speech-synthesis model, could be integrated into Orbit, Bakken & Bæck’s natural-language platform. I also created, annotated and cleaned training data for Norwegian pronunciation.
Stack
Python, TensorFlow, NumPy, SciPy
02
About
I’m a backend engineer who likes owning a system end to end: the data model, the APIs, the infrastructure it runs on and the pipeline that ships it.
Over more than ten years I’ve worked in early-stage startups, a design and technology studio and a global investment bank.
I studied Software Engineering at Edinburgh Napier University, spent a year at the University of Utah, and took an MSc in Artificial Intelligence at the University of Edinburgh, specialising in intelligent robotics. A summer internship in Trondheim first drew me to Norway. I moved to Oslo in 2016, spent six years building products at Bakken & Bæck, and have been in London since 2022.
Fig. 4 One system, end to end.
The route so far
Edinburgh2010 – 2016BEng, first startup job, MSc
Trondheim2012Summer internship
Salt Lake City2012 – 2013Study-abroad year
Oslo2016 – 2022Bakken & Bæck
London2022 – presentBank of America
How I work
1
Own it end to end.
Architecture, APIs, infrastructure and delivery are one job, not four.
2
Pragmatic by default.
Ship early, iterate fast, and spend complexity only where it pays.
3
Built to last.
Maintainability, extensibility, robustness and scalability are design inputs, not clean-up tasks.
4
Automate the path to production.
Tests, CI/CD and infrastructure-as-code, so that releasing is routine.
03
Experience
Sep 2022 – present
London, UK
Bank of America
Senior Front Office Developer (VP), XVA Technology
I own and evolve Python-based trading tools for the XVA desk, working directly with trading, quant, finance and risk teams in London and New York.
Design and maintain high-performance REST APIs behind internal trading workflows and analytics.
Improve the reliability of data pipelines and reporting systems, keeping daily reconciliation of trading data consistent.
Build React frontends for desk tools, backed by Python (Flask) services.
Work closely with traders and quants to deliver production-grade tools on fast iteration cycles.
Deploy and manage services through CI/CD pipelines (Jenkins), with automated unit and integration testing.
Stack
Python, Flask, Tornado, React, Jenkins
Dec 2016 – Aug 2022
Oslo, Norway
Bakken & Bæck
Senior Software Engineer
Bakken & Bæck is a design and technology studio, founded in Oslo in 2011.
I led backend development across multiple zero-to-one products, owning architecture, scaling and production delivery, and worked on in-housemachine-learning projects with TensorFlow and PyTorch.
I mined web data with Go, built a web application with Node.js and Angular 2, built a desktop AWS client in C++ with Qt, and used Unreal Engine for 3D architectural visualisation.
Stack
Go, Node.js, Angular 2, AWS, C++ / Qt, Unreal Engine
May 2014 – Jul 2015
Edinburgh, UK
Make it Social
Backend Developer
Make it Social was an Edinburgh startup building group-booking and group-payment software. I built APIs and web services in Python (Django) and Go to meet new clients’ needs, and built and maintained the AngularJS app.
Stack
Python, Django, Go, AngularJS
Earlier, 2012 – 2014
Edinburgh Napier University, for Royal Bank of Scotland
Coding Trainer Nov 2013 – Jan 2014 · Edinburgh
Part of a university team that taught around 30 Royal Bank of Scotland staff to code in Java and JavaScript.
Edinburgh Napier University
Database Systems Demonstrator Sep – Dec 2013 · Edinburgh
Supported second- and third-year students in practical sessions.
University of Utah
Web Developer (Intern) Mar – May 2013 · Salt Lake City
Maintained the website of the university’s MUSE project and added new features, in PHP.
PoleVault Media
Web Developer Sep 2012 – Feb 2013 · Salt Lake City
Built WordPress sites with custom JavaScript interactions.
SiT (Studentsamskipnaden i Trondheim)
Web Developer May – Jul 2012 · Trondheim
Planned the mobile website for Trondheim’s student welfare organisation.
Volunteering
IAESTE
Volunteer, IT support Since 2012
IAESTE arranges paid technical work placements abroad for science and engineering students. I started as a local committee member and IT officer in Edinburgh, looking after incoming trainees, and I’m now part of the volunteer team that supports its international Exchange Platform. iaeste.org
04
Toolbox
Languages
Python expert
Go
TypeScript
JavaScript
SQL
Backend
Flask
Tornado
REST APIs
SQLAlchemy
Pandas
NumPy
Frontend
React
Data
PostgreSQL
MySQL
Cloud and infrastructure
AWS (EC2, RDS, S3, VPC, Route 53, CloudFormation)
Kubernetes
Docker
Nginx
Delivery
Jenkins
Git
CI/CD
Test-driven development
Agile
DevOps
Also worked with
TensorFlow
PyTorch
Django
Node.js
AngularJS
C++ / Qt
OpenCV
MATLAB
Java
C#
PHP
The stack through the years
Where I was, and what I was mostly writing.
2010 – 2014
University, Edinburgh and Salt Lake City
PHP
Java
C#
C++ / Qt
JavaScript
OpenCV
2014 – 2016
Edinburgh startups and MSc
Python (Django)
Go
AngularJS
Node.js
MATLAB
2016 – 2022
Bakken & Bæck, Oslo
Python (Tornado)
PostgreSQL
MySQL
AWS
Kubernetes
Jenkins
React
TensorFlow
PyTorch
2022 – present
Bank of America, London
Python
Flask
Tornado
React
Jenkins
05
Research and publications
BEng honours project
Edinburgh Napier University, 2014
A laser pointer as a touchscreen
My honours project explored using a laser pointer as a mouse: a webcam watches a projected screen, and image processing with OpenCV turns the laser dot into pointer movement and clicks. It became a full paper at ICEIS 2015 in Barcelona, and an extended version was one of 30 papers selected, from 327 submissions, for Springer’s Lecture Notes in Business Information Processing.
Conference paper
Aizeboje, J. and Peng, T. (2015). An Approach to using a Laser Pointer as a Mouse. Proceedings of the 17th International Conference on Enterprise Information Systems (ICEIS 2015), Barcelona, pp. 543–552. SciTePress.
Aizeboje, J. and Peng, T. (2015). An Approach to a Laser-Touchscreen System. Enterprise Information Systems: ICEIS 2015, Revised Selected Papers. Lecture Notes in Business Information Processing, vol. 241, pp. 476–495. Springer.
Recognising ceilidh dances from an overhead camera
Recognising which stage a ceilidh-style group dance has reached, from overhead video. I designed a modified linear-chain conditional random field for ordered state transitions and compared it with the standard model. Given labelled positions of the ten dancers, the modified model identified each stage of two dances with 96.85% and 94.42% accuracy (within a five-frame tolerance), on average 6.8 percentage points better than the standard model. The dissertation is the reference that users of the Edinburgh Ceilidh Overhead Video Data are asked to cite, and the dataset was later used in an ECCV 2022 paper.
Dissertation
Aizeboje, J. (2016). Ceilidh dance recognition from an overhead camera. MSc dissertation, School of Informatics, University of Edinburgh. Supervised by Prof. Robert Fisher.