Founder operating system

Tony Qiu

Founder & CEO of Aurem Group ABEngineering × Finance × Execution

I build control and evidence systems for consequential AI-agent actions while completing parallel technical and finance degrees at KTH and SSE.

A structured map connecting engineering, finance, and company building
02parallel degree tracks
03live products in focus
CEOAurem Group AB

01 / Founder

Operating across three systems.

The hardest products sit where technical possibility meets financial consequence. I work across both—turning ambiguous risk into software, evidence, and decisions people can inspect.

Tony Qiu in Stockholm
Stockholm, Sweden · Founder / engineer / student
SYS / 01

Engineering

Build the real system.

Product architecture, full-stack implementation, data systems, and interactive technical proof.

SYS / 02

Finance

Model the consequence.

Risk, incentives, decision quality, and the commercial logic that determines whether software matters.

SYS / 03

Company building

Ship toward evidence.

Positioning, product judgment, design-partner readiness, and measurable next decisions.

02 / Company

Founder & CEODesign-partner beta

Building the authority layer for consequential AI actions.

Aurem is an authority and evidence layer for high-risk AI-agent actions—binding identity, action rights, runtime limits, approvals, and reconstructable evidence.

My role. I lead the product, engineering, and commercial direction: from the Agent Passport and runtime-control model to the public product experience and design-partner path.

  • Agent Passports
  • Runtime authority
  • Human approvals
  • Evidence chains
Aurem authority network on the public control-room homepage

03 / Selected work

Products first. Proof attached.

Selected for founder relevance, technical difficulty, and evidence. Contribution boundaries are explicit; live and source links are shown only where they exist.

Aurem control-room homepage with an authority network visualization

01 / Founder companyFounder-led

Aurem Risk

Authority, runtime control, and evidence infrastructure for enterprise AI agents.

Problem
As AI agents gain access to tools, data, and spend, organizations need to know who authorized an action, which boundary governed it, and how to reconstruct what happened.

Exact contribution

  • Founded the company and set the product thesis.
  • Led the control model spanning identity, policy, approvals, runtime limits, and evidence.
  • Built and shipped the product, public site, and design-partner readiness system.
  • TypeScript
  • React
  • Three.js
  • PostgreSQL
  • Google Cloud
Bevisa homepage showing an interactive tender evidence field

02 / Sole-author productSole author

Bevisa

An evidence and compliance operating concept for Swedish public tenders.

Problem
Tender teams can lose requirement coverage when evidence lives across documents, owners, and expiry dates.

Exact contribution

  • Designed the product narrative, evidence model, and bilingual interface.
  • Built the interactive evidence field and progressive WebGL fallbacks.
  • Added content, accessibility, bundle, browser, and deployment gates.
  • React
  • TypeScript
  • React Three Fiber
  • Playwright
Capasight editorial homepage asking whether a strategy can be staffed

03 / Sole-author productSole author

Capasight

A pre-commitment talent-feasibility gate for Nordic workforce decisions.

Problem
Strategy can be approved before leadership knows whether critical roles can be supplied and attracted in the chosen market.

Exact contribution

  • Defined the narrow decision-gate positioning and evidence flow.
  • Built the static-first Astro experience and editorial visual system.
  • Encoded source, wording, accessibility, and safe conversion-link controls into CI.
  • Astro
  • TypeScript
  • Playwright
  • Axe

Proprietary internship work

Computer vision return pipeline

Computer-vision and software pipelines for e-commerce return recognition.

Sole-author system

Course planning & allocation system

A transactional Java and PostgreSQL system for course planning, staffing, and salary versioning.

Team analysis · Python lead

Crowdfunding data analysis

A collaborative analysis of campaign and contributor data, with Tony owning the Python work.

04 / Parallel tracks

Engineering and finance, in parallel.

Two independent degree tracks create the working vocabulary behind the products: how systems behave, how risk compounds, and how decisions earn confidence.

Track 01

KTH Royal Institute of Technology

Information & Communication Technology

Engineering track focused on software, systems, mathematics, data, and computing infrastructure.

2024 — present

Track 02

Stockholm School of Economics

Business & Economics

Finance and business track focused on markets, decision-making, strategy, and commercial judgment.

2024 — present

KTH / technical systemsSSE / financial systems

05 / Experience

Roles closest to execution.

A selective timeline: company building, applied technical work, and institutional representation.

Current

Founder & CEO

Aurem Group AB

Leading product, engineering, and commercial direction for agent-assurance infrastructure.

Current

International student ambassador

KTH EECS

Representing KTH and EECS in international student outreach and recruitment.

2024 — 2025

Computer vision & software engineering intern

Pango

Worked on computer-vision models and operational software pipelines for e-commerce returns.

06 / Proof layer

A short path from claim to source.

Public evidence is one click away. Private transcripts, contracts, source repositories, and confidential work stay private.

Evidence checked on 2026-07-17. Product and repository state can change after launch.

07 / Contact

For consequential systems and ambitious builds.

Founder conversations, technical collaboration, and thoughtful introductions are welcome.