Its ring is pressed as vinyl – four tracks, one per language.
Read the case →
1 / 8
Max Sushchuk AI Transformation Lead · Kyiv
I build the parts of a company that don't go home at night.
Some run for clients, some run my own week. Either way, the best hour I get is the one where you explain a problem I have not seen before.
Write about what is not working, and we'll go looking for the sore spots and the growth areas together. More on that.
Hover a world to read it. Select Opportunity radar to follow it into its case.Tap a world to fly in. Opportunity radar takes you into its case.
How to read a world
Every world's anatomy is counted from its public case file. Select a feature to see it across the whole system.
- The sunThe workspace the rest runs on – 8 streamers for 8 hosts, 38 bright cells for 38 services.
- VinylContent operator wears its ring as a record – its signature. The four tracks are its four languages.
No mark means the case gives no number for it. Nothing is invented.
The numbers behind each world
| World | Moons · agents | Ring · divisions | Lights · volume | Field · tests | Status |
|---|---|---|---|---|---|
| Swarm workspace the sun | 8 hosts · 38 servicesstreamers and bright cells · metric row | – | In progress61 days build · ongoing | ||
| Content operator | 1 agentframe | 4 languagesmetric row · pressed as vinyl | 17 articles shippedmetric row | – | Finished~3 months build · ongoing tuning · two writers |
| Growth hacker | 3 agentsframe | – | 14 measured outcomesevidence list | 222 teststerminal scene | In progress23 days to a live loop · in production · rollout ongoing |
| AI-Ops kit | 3 subagentsmetric row | 6 systemsmetric row | – | – | Finished~3 weeks solo build · shipped to the ops team |
| Krapka | – | 8 languagesmetric row | 1,632 pieces of prosecase text · 68 × 3 × 8 | 109 testsmetric row | Finished9 months build · live on real payments |
| Opportunity radar | – | 74 sourcesmetric row | 9,114 read · 135 forwardmetric row | – | In progress~5 weeks build · 319 commits · runs every morning · in use by the product team |
| AI Recruiter | 1 intake agentframe | – | – | 1,190 testsmetric row | Finished6 days to first ship · two months hardening under live traffic |
| Sales-intel agent | 4 agentsmetric row | – | ~1,000 records / weekmetric row | – | Finished~8 months iteration · in production · architecture stable |
Counts come from the public case files – their metric rows, frames and evidence lists. Light counts use a square-root scale (three lights per square root of the volume) so a thousand records fit on a small world; field loops use a log scale. Where lights and cells sit on a world, the moons' sizes and periods, and the orbits' small eccentricities and inclinations are illustrative. Orbit tiers, angles, tilt, spin, size and atmosphere hue come from the site's own art and editorial records.
-
Swarm workspace
My personal multi-agent workspace – the operational foundation everything else runs on. A long-context AI agent as operating partner against a version-controlled homelab-ops repo; a hardened bastion as the blast-radius boundary; the rails the proactive-AIOps turn will run on.
-
AI-Ops kit
A Claude Code skill catalogue with subagents that turns a multi-system operations investigation into a single conversation. Built solo in about three weeks; 15 skills and 3 subagents, shipped to the operations team.
-
Sales-intel agent
A chat-triggered agent that finds B2B contact intelligence for a sales team. Salespeople ask in plain English; the agent returns enriched records via a spreadsheet. 7 months in production, ~1,000 records per week.
-
Opportunity radar
Every morning from 05:00 UTC the radar reads the whole market a B2B company sells into – 66 live sources out of 74, its own customers beside tenders, regulators, competitors, research and the trade press of every industry it serves – weighs each signal against its own history and has a tray of opportunity cards ready before the working day starts.
-
AI Recruiter
End-to-end recruitment automation, soup to nuts: a marketing landing page, a chat-app intake agent that interviews candidates, an admin SPA for recruiters, and a SQL backend tying it together. Live in production, rebuilt so that every claim the agent makes is verified against tool evidence. 1,190 automated tests cover behaviour.
-
Content operator
Turns a content brief into a published article, in 4 languages. Stages: normalize → validate → render → publish. About three months of work; used daily by 2 writers.
-
Krapka
A typing trainer for reading foreign-language literature – type real stories at your level, in the alphabet you are trying to make automatic. Built and operated solo: 68 public-domain stories, each with three graded retellings glossed sentence-by-sentence into seven more languages.
-
Growth hacker
A closed growth loop run by three agents and one operating surface: the first finds where AI answer engines cite a competitor instead of you, the second drafts the page fix, the third measures whether the citation came back – and feeds that into the next draft.
what people bring me
-
Tenders, reports and trade news about your market pile up faster than your team can read them.
A radar reads 66 sources every morning and has opportunity cards on the tray before the working day starts – about 180 analyst-hours a week. Opportunity radar →
-
Sales spends hours working out whom to write to.
One line in Slack, and a minute or three later a table of contacts – deduplicated, with honest sources. About a thousand a week for a sales team of five. Sales-intel agent →
-
The first conversation with every candidate eats the recruiter's day.
An agent holds it in chat in three to five minutes, around the clock. Code makes the decisions, so the agent cannot promise what is not there. AI Recruiter →
-
A brief takes most of a working day to become an article.
An agent takes it to a formatted Google Doc in 20 minutes and translates it into four languages. Two writers use it every day. Content operator →
-
More growth ideas than hands to test them.
Three agents propose changes, prepare them for release and measure the result. A person spends fifteen minutes a morning on the decisions. Growth hacker →
-
The answer to an operations question is spread across six systems.
Claude Code skills and three research subagents turn it into one conversation – with clear rules, every step logged, safe enough to hand to the whole team. Built in three weeks. AI-Ops kit →
how it starts
- Write about what is not working, in your own words, and what you have already tried. A rough note is fine.
- Say whether you want an opinion or a pair of hands. An opinion is what I actually think you should do – including "you do not need me for this", when that is the answer. A pair of hands means I build it; the cases above are what that has looked like.
- Let's meet. I am in Kyiv and would rather talk in person: a conversation over coffee usually gets further than a call.
what stays between us
No client is named on this site. The cases show the architecture, the decisions and the numbers – never the company, and never its market closely enough to recognise it. The one product here under its real name is my own, Krapka.