01 · Product, 0 → 1
Skipy, the v1 of Traast
Built Skipy, our MVP, solo and put it in production with real users. Our CTO later rebuilt it as Traast.
Read the case study →Fig. 1 - a demo of Skipy, the v1 I built, fictional data
I plan products, ship v1, and hand it to engineers.
Then build the machine that sells it.
Wrote the PRD for Skipy, the MVP that became Traast, built it myself, and handed our CTO a v1 that already had users in production. Then built the outbound and sales engines, the signal radar, and the skills that ran our sales and GTM.
Fig. 0 - the operator
24/7
automated signal monitoring
1,183
commits
26+
agent skills
25+
tools integrated
Selected work
02 · GTM engineering
The Lead Machine
The whole outbound motion as one system, sourcing to reading replies. One operator runs what a full sales team would.
Read the case study →Fig. 2 - a demo I built: one run of the machine, fictional data
03 · GTM engineering
Signal Scout
Reads founder groups on Telegram, classifies each post with Claude, alerts me first. Only leads worth acting on.
Read the case study →Fig. 3 - a demo I built: one run of the machine, fictional data
04 · Personal systems
A personal assistant for founders and managers
Watches my calendar, tasks and CRM for habits I repeat, nudges me at the moment they recur.
Read the case study →Fig. 4 - a demo I built: one run of the machine, fictional data
And more
- This siteDesigned and built by me with an AI coding agent. No template, no agency, no designer. The four interactive demos, the galaxy at the bottom of this page (a 3D engine in about 3,600 lines, no graphics libraries, lit by the music's real spectrum) and the Ask box below are all part of it. 116 commits in its first 10 days.
- Models behind the decisionsAlongside the working systems, I built business plans, budgets and scenario forecasts in spreadsheets. Assumptions and calculated outputs stay explicit, so the model can support a decision and be revised when the inputs change.
- A library of agent skills →26+ reusable skills across sales, marketing, product research and day-to-day ops: LinkedIn connector, outreach status auditor, warm-intro path builder, CoreSignal-to-cold-email generator, meeting follow-ups, pitch-deck builder, competitor research.
- Every tool, mastered and wired inI go deep on each tool, past the login, then connect it to the rest through its API, webhooks or a custom MCP server. CRMs, inboxes, chat, data sources and AI models end up as one working system.
- Instinct
- Muse
- Grok Bot
- Apollo
- HubSpot
- Claude API
- OpenAI API
- Slack
- Claude Code
- Claude Design
- Codex
- Lovable
- Base44
- GitHub
- Notion
- Todoist
- Telegram
- Gmail
- Google Sheets
- Google Calendar
- Google Drive
- CoreSignal
- Instantly
- MeetAlfred
- Custom MCP server
Ask my work a question
Live questions get an AI answer grounded only in this site's own project data, never the model's free-form guessing.
answered from this site's data
That one's better answered live.
Book 30 minoranov.con@gmail.comLet's build something.
I'm interested in working with people with real ambition, building great things. Got a role, a project, or anything else that fits? I'd love to hear about it.
Book 30 minoranov.con@gmail.comLinkedInGitHub