Operator-investor working across AI x crypto systems, digital assets, institutional workflows, and privacy-first agent systems.
Founder at SAIBA. Building useful systems with memory, verification, and human judgment.
The pinned shelf is the front door. The wider archive below is the map.
- SAIBA Agency OS - live memory, Hermes cards, scoped agent work, review gates, and execution receipts.
- Customer Event Flywheel - a public spec for turning customer signals into shared memory and safe next actions.
- sai-platform - MCP-first, eval-driven starter architecture for reusable agent systems.
- UserOwned.AI - prior AI x crypto intelligence work from the NEARWEEK lane.
- NEARWEEK automation - content and intelligence automation for fast-moving open-web ecosystems.
- This profile - the public operator map across AI x crypto, digital assets, governance, and privacy-first agents.
SAIBA is a privacy-first operating layer for expert teams that need agents to help without turning customer work into a black box.
The loop is deliberately simple:
customer signal -> shared memory -> human decision -> scoped agent work -> receipt
That means discovery calls, Slack handovers, CRM notes, GitHub issues, delivery reviews, and customer objections become structured context. Agents can help with the work, but risky artifacts stay behind human approval.
The public content system is built around source-tethered posts: real releases, PRs, demos, and customer-safe receipts become drafts for the right person or company account. The point is not to flood channels. The point is to make real work easier to understand from the outside.
The technical workbench is MCP-first, testable, and receipt-oriented. I care about systems where the agent can explain what it saw, what it changed, why it changed it, and which human approved the public edge.
This archive mixes current systems, prototypes, and older ecosystem artifacts. Together, they show the throughline: crypto, governance, intelligence, distribution, and now privacy-first agent workflows for institutions.
| Project | What it shows | Read it as |
|---|---|---|
| saiba-agency-os | Human-approved agent orchestration for expert-service teams | Current SAIBA operating thesis |
| customer-event-flywheel | Customer signals routed into memory, decisions, and safe action | Audit trail and memory spine |
| sai-platform | MCP-first agent starter architecture and eval-driven development | Agent systems workbench |
| Kisgus | This profile surface and public positioning shelf | Public operator map |
| Project | What it shows | Read it as |
|---|---|---|
| userowned-ai | Prior AI x crypto intelligence and ecosystem signal work | NEARWEEK intelligence lane |
| userowned.ai | Public-facing AI x crypto ecosystem intelligence platform | Media and intelligence edge |
| userowned-ai-website | Website layer for AI x crypto intelligence | Distribution surface |
| nearweek-automated-news-sourcing | Human-approved sourcing and publishing workflows for NEARWEEK | Source-to-public workflow |
| nearweek-intelligence-unified | Unified intelligence interface for ecosystem monitoring | Prototype integration layer |
| intelligence-network | Verifiable crypto news and agent-assisted ecosystem monitoring | Intelligence architecture |
| nearweek-telegram-bot | Telegram distribution for NEAR and crypto updates | Distribution automation |
| userowned-ai-buffer-airtable | Buffer, Zapier, and Airtable workflow for content operations | Content ops plumbing |
| Project | What it shows | Read it as |
|---|---|---|
| multisig-tool | Practical wallet and treasury coordination tooling | Asset governance context |
| proposals | House of Stake proposal context | Governance participation |
| ui.grants | Grant application tooling for DAOs on NEAR | DAO operations context |
| admin.news | Earlier admin and publishing infrastructure | Media operations history |
| EIPs | Ethereum standards ecosystem exposure | Standards literacy |
| ethereum-org-website | Ethereum public-goods documentation context | Ecosystem literacy |
| celo-monorepo | Celo ecosystem exposure | Protocol ecosystem context |
| GECO | Gnosis Ecosystem Fund context | Ecosystem investment history |
| Linkdrop-Relayer | Link-based token distribution infrastructure | Crypto onboarding infrastructure |
| eth2-app | Early token transfer and wallet interaction experiment | Consumer crypto UX history |
| CryptoLife | Hackathon-era crypto participation | Early ecosystem participation |
| trustlines-network.github.io | Early decentralized credit network context | Network and community currency history |
| Project | What it shows | Read it as |
|---|---|---|
| nft-dropshipping-shop | Commerce, Shopify, and NFT ownership certificates | Commerce and ownership test |
| spotify-app | Web app authentication and media controls | App-building experiment |
| Cheria | Portfolio/client-style web work | Older web surface |
| Kisgus.github.io | Earlier personal publishing surface | Legacy profile surface |
| crypto-chrismas | Early consumer crypto campaign experiment | Prior experiment |
| cryptoxmas.xyz-images | Campaign media archive for a crypto giving experiment | Prior media artifact |
The broader thread in my work is how AI, crypto, and digital assets become useful inside institutions without losing privacy, agency, or trust.
I have worked across open-web products, protocol ecosystems, public-sector blockchain research, AI x crypto intelligence, and digital asset investment communities. That includes NEARWEEK, Linkdrop, Trustlines, brainbot/Raiden, community currencies, public-sector research, and advisory work around institutional adoption.
My current focus combines four lanes:
- privacy-first agentic workflows through SAIBA,
- digital asset infrastructure and investment through Stormshield ApS,
- ecosystem intelligence and distribution through NEARWEEK,
- public-sector and institutional governance around AI x crypto.
- Proof before narrative.
- Human approval before customer-visible sends.
- Local-first when the data is sensitive.
- Receipts for important agent work.
- Strategy that can become software.
- Public writing tied to shipped work, not generic commentary.
I am most useful to founders, funds, family offices, and institutions when the question sits between technology, trust, and execution:
- turning a thesis into a working product or operating system,
- understanding AI x crypto infrastructure before it becomes obvious,
- building agent workflows that fit real organizational constraints,
- making content and intelligence systems source-tethered and reviewable,
- translating policy, market structure, and technical reality into decisions.
How do we make AI, crypto, and digital assets useful enough for institutions
without losing privacy, agency, or trust?
- Website: saiba.dk
- GitHub: Kisgus
- LinkedIn: Gustav Friis