hey, I'm erikrasin.io 👋
Erik Rasin
Finance × AI Engineering.
A decade inside banks and insurers, working cash, liquidity, and treasury, taught me how financial institutions actually run. I now build applied AI: agents, retrieval, and workflow automation for research and analysis, focused on where it creates the most value rather than where it is merely possible. I ship these end to end, from data pipeline to interface, using full-stack and cloud engineering.
About
I build applied AI systems, with a focus on shipping something useful, not just impressive. My path here is unusual: twelve years inside regulated financial institutions, across treasury, private banking, and IT consulting, followed by two years building and operating AI products independently. That combination is the lens I bring to this work: I understand what finance and other data-heavy businesses actually need from software, and I know how to build it, from agents and retrieval systems to the full-stack and cloud infrastructure underneath them. This site shares my projects, technical writing, and experiments in applied AI, finance systems, and infrastructure.
Focus Areas
Applied AI & Agents
Designing agents, retrieval, and tool-use pipelines that automate research, analysis, and decision support, built to run in production, not just in a demo.
Finance & Institutional Workflows
Twelve years inside cash, liquidity, treasury, and private banking at major banks and insurers. I know what institutional workflows actually require, and I look for where AI creates the most value, not just where it is possible.
Full-Stack Product Engineering
Building production web applications end to end with React, Next.js, TypeScript, and PostgreSQL.
Cloud, Data & Reliability Engineering
CI/CD pipelines, infrastructure as code, and containerized deployments on AWS and Azure, with the audit trails and evaluation harnesses that keep AI systems trustworthy running underneath.
Reliability & Governance
Capable models are the easy part now. Making them reliable enough to trust with real financial and operational decisions is the harder, and more interesting, problem. I write about how to evaluate agents, test them under real conditions, and report results honestly: the same rigor I would expect from any system handling other people's money.
Navigating the Future of Agentic AI Evaluations: From Static Prompts to Dynamic Sandboxes
An end-to-end guide to evaluating agentic AI: the capability-reliability gap, the four pillars of evals, dynamic sandbox testing, and the governance reforms reshaping how benchmarks are reported.
Projects
Products and experiments, built end to end, from data pipeline to interface.
AI-native trading and research interface: real-time market data, interactive charts, and conversational analysis in one workflow.
AI-assisted follow-up and outreach workflow that helps small B2B service businesses turn more inbound inquiries into booked calls.
Agent-driven web research: search, synthesis, and summarization pipelines built on top of LLMs.
Summarization tool that extracts and condenses long-form content into something usable in minutes.
Writing
Technical notes on applied AI, finance systems, and the infrastructure underneath them, grouped by theme.
AI & Machine Learning
Cloud, DevOps & Security
Working Style
I work best in small, focused teams where there is a hard technical problem and room to go deep. Comfortable as a solo builder or embedded in a larger group as a technical lead. Remote-first by default, based in Zurich, and open to relocating for the right role.
Contact
If you are building AI tools for research, analysis, or workflow automation, especially in finance or another data-heavy business, and want to find where it creates the most value, I am happy to talk. A 15-minute call is usually enough to figure out if there is a fit.
Thanks for stopping by. 🚀