Projects / work in progress
Useful systems.
Open questions.
Some projects help a business run better. Others start with curiosity. Here is what I’m working on, and what the evidence actually says.
AI & operations / current
Orvus
I co-founded Orvus and build AI systems and integrations. My focus includes customer-service automation, SEO and lifecycle systems, connecting marketing strategy with implementation.
Public first-party case summaries describe support workflows and SEO work. Support actions, ticket closures and fully automated conversations are different measures. Search impressions, clicks and realized revenue are different outcomes.
Reviewed 15 September 2026. Detailed measurement windows and attribution are needed before I would reproduce the public percentages as personal performance claims.
Read Orvus’s case summaries ↗Customer service / systems integration
Customer-Service Brain
My customer-service work at Orvus connects support conversations with business knowledge, order information and shipping workflows. It comes from the practical work of running ecommerce operations: getting the relevant context into one workflow and keeping people involved where judgment is needed.
Automated actions, approval-dependent work and human handoffs are different parts of that workflow. Action counts are not counts of customers served without a person, and model usage is not the total cost of support.
Sign in to Customer-Service Brain ↗
Access is for authorized users. Historical performance figures will be referenced here once the dated case study and its source evidence are published.
Built in public / initial trial
A network of specialist publications.
At Orvus, I’m developing a media network with distinct editorial identities supported by shared research, publishing and measurement systems. Ten blogs were the initial trial. Potential expansion to hundreds depends on reader demand, quality and commercial results; it is not today’s scale.
10 sites. 173 impressions. 0 clicks.
This baseline covers 31 August to 13 September 2026, collected on 14 September 2026 from Search Console. Impressions mean appearances in search results, not visits. The report contains no revenue measurement. Search Console data can lag and query reporting can be incomplete.
Source: my first-party operational Search Console report, with all ten properties reporting successfully. This is an early baseline, not an independently audited growth case study.
Archaeological intelligence / ongoing research
Tabula
I’m building Tabula with my brother, Jakob. It brings historical literature, archival maps and LiDAR terrain data together to identify places worth investigating and make the reasoning behind each lead easier to inspect.
The protected workspace brings captured sources, map layers, terrain observations and inspectable site arguments together. More extensive automated source reading and site-specific interpretation remain in development.
A prediction is a starting point for investigation, not a confirmed discovery. Documentary research, field checks and appropriate expert assessment determine what it means.
Project introduction, 15 September 2026. Public explanation and sign-in are separate from the protected research workspace.
Have something worth building?
I’m open to client work, strategic partnerships and senior marketing and growth opportunities.
Let’s talk ↗