Hey, I’m Jaakko 👋
I build software across backend, infrastructure, and applied AI. I care about simple systems, clean interfaces, and measurable quality.
Final-year Master’s student graduating in spring of 2026 in Software Engineering from LUT-University, focused on DevOps, generative AI systems, and web development. My work sits at the intersection of building reliable software and running it well: shipping services that are observable, testable, and maintainable, while understanding the infrastructure they live on.
I’m currently developing a developer-facing generative AI testing and evaluation service as my Master’s thesis in collaboration with Telia. The goal is to make LLM-based features measurable and dependable by evaluating groundedness/faithfulness, tracking regressions, and supporting modular, scalable integration into real systems. Earlier academic research on AI and software development has seen 600+ downloads from my bachelor’s thesis.
Alongside research, I run a Kubernetes + Linux homelab on Proxmox to practice production-style operations: clustering, networking, service deployment, and day-to-day reliability work. This environment is where I turn concepts into repeatable workflows and build intuition for how systems behave under change.
Right now I’m deepening practical skills in GitOps, AWS, and Infrastructure as Code, while continuing to build on-prem AI solutions and full-stack projects. If you want details, the Projects section on my LinkedIn page covers the concrete implementations and the technical choices behind them.
If I’m not working on a project, I’m usually reading, writing, or tuning my homelab. I write here.
This site is still work-in-progress, some content may be subject to change or temporary.
Contact: jaakko.lipponen01@gmail.com · LinkedIn · GitHub
PORTFOLIO
Developer-facing backend service for evaluating LLM output and making developing Generative AI systems more testable with DeepEval.
The system is designed to be modular and scalable. Currently working on this as a part of my master's thesis, in collaboration with Telia Finland.
Personal 4-node proxmox cluster homelab for learning DevOps skills: local LLM models, high availability hosting, kubernetes, and KVM virtualization.
Public access to locally hosted services is achieved via a wireguard reverse proxy with a rented VPS.
Many other smaller projects varying from ESP32 controlled custom smart light solutions to AI software patchnote summarizers, LLM chatbots on Discord, self hosted web application ci/cd pipelines, ...
A deep-dive bachelor's thesis into AI-driven software development practises and LLM model viability.
Graded 4 out of 5 with 600+ downloads as of Q4 2024. I aim to ship small and efficient, measure, iterate, automate, and keep the surface area minimal when developing software.
PREVIOUSLY
- Technical specialist & masters thesis worker · Telia Finland2023–2026 · Microservices, evaluation, RAG, infraCurrently building an LLM evaluation platform and internal tooling for quality control of AI outputs. Also worked in key corporate networking solutions fault management, service assurance, network security operations.
- M.Sc. Software Engineering · LUT-Univesity2021–2026 · Software Development, DevOps, Architectues & ModelsCoursework and projects spanning distributes software systems, Linux, databases, and sustainable DevOps practises.
USER MANUAL
A short playbook on how I work: preferences, strengths, and how to collaborate effectively. W.I.P.
More about me →