Language & reasoning
Giving language models context, tools, and a clear purpose.
- LLMs & fine-tuning
- Retrieval-augmented generation
- Prompt & context engineering
I build AI systems that move beyond possibility.
Thoughtfully engineered. Grounded in the real world.
Made to be useful.

Language models· Intelligent agents· Reliable systems· Human impact
I connect the possibilities of AI
with the realities of building for people.
A collection of illustrative project concepts. Case studies are examples, not claims of shipped work.
I’m Elsa, an AI engineer drawn to the space between what a model can do and what people actually need.
I approach AI as a systems problem: the data, the model, the evaluation, and the experience all have to work together. I care as much about the quiet details — a useful fallback, a clear explanation, a reliable deployment — as the breakthrough moment.
My north star is simple: build things that earn people’s trust, and keep asking better questions along the way.
From the first experiment
to a system you can depend on.
Giving language models context, tools, and a clear purpose.
Building systems that act with intention, within thoughtful boundaries.
Measuring what matters. Understanding where a system falls short.
Bridging experimentation and the realities of production.
Good AI engineering is a practice, not a single model choice. Here’s how I move from an open question to a dependable system.
A working philosophy. A verified career timeline will be added here.
Start with the person, the problem, and what success would actually look like. A clear question is the most useful tool.
Build a small baseline, test assumptions, and evaluate honestly. Let evidence decide what earns a place in the system.
Design for latency, cost, failure, and change. Keep the system observable and leave room for human judgment.
Have a meaningful problem, an ambitious idea,
or a different way of seeing things?
I’d love to hear about it.