53 lines
8.2 KiB
Markdown
53 lines
8.2 KiB
Markdown
# Eugene
|
||
|
||
#entity #person
|
||
|
||
## Summary
|
||
|
||
The interviewer in [[2026-07-06-sebastian-interview-ai-and-software-engineering]]: a hands-on computer-vision & embedded/firmware developer, content creator, and builder of a custom Claude Code "harness." His lens is individual leverage and getting closer to the client's real problem. He appears likely to be the owner of this vault (see Contradictions / Uncertainty).
|
||
|
||
## Current Understanding
|
||
|
||
Eugene's signature positions:
|
||
|
||
- **"Build your own harness."** Every developer should build a personal harness on top of Claude Code; knowing every detail makes it far more effective. He demoed his own: a Telegram-like UI, one agent per project, inter-agent messaging, per-agent memory, a "done thinking" signal so you don't babysit the console. Tools he references: **Conductor** (git-worktree-per-chat, automated PRs/merges). See [[future-of-engineering-work]].
|
||
- **His stated unsolved problem (stuck ~6 months): how to build a professional network.** Sebastian's "connections are everything" answer is aimed squarely at this — see [[relationships-as-moat]]. Eugene had been investing in LinkedIn/articles; the interview's takeaway for him is to reallocate to recurring in-person events.
|
||
- **A target for the networking itself (added 2026-07-22).** [[partnerships]] reframes "build a network" — vague, and stuck for 6 months — as **recruit a specific partner archetype**: identify who already holds trust with his target buyers (for CV/embedded work, plausibly industrial-equipment vendors and machine-builder integrators — untested inference), find the events *they* attend, win individuals. This composes with both diagnoses below rather than competing: it supplies the criterion that makes showing up purposeful, and a partner intro partially substitutes for self-articulation because the partner does the vouching. Same-author, anecdote-grade source ([[dan-martell]]).
|
||
- **A concrete day-one outbound move (added 2026-07-23).** [[2026-07-23-make-my-first-100k-in-month]] supplies the cheapest test of the network blocker: mine existing phone contacts and **"ask past the person"** ("do you know anyone with this problem?") — warm, referral-shaped intros that require no past clients and no publishing ([[referrals]]). Costs an afternoon; would also produce the first real data on whether his network is thin or merely unasked.
|
||
- **A stage/format mismatch worth naming (added 2026-07-23).** The same source's recommended shape — ~100 SMB customers × ~$1K/mo, closed by chat-DM and cold calls — fits CV/embedded work poorly: his natural deals look like few × $10K+ (the "enterprise-lite" shape Martell dislikes at $0 but serves with [[partnerships]] at scale). Applying the blueprint to him likely means keeping its *sequence* (offer → pre-sell → build; phone-mining outbound) while rejecting its price×count sweet spot. Vault inference, untested.
|
||
- **A pre-build validation checklist he can run this week (added 2026-07-26).** [[sell-before-build]] ([[2026-07-26-main-principle-of-successful-business]]) turns the vault's validate-first rule into concrete moves for his situation: name the smallest group with the most acute pain **by name** (10 concrete industrial-equipment/machine-builder prospects), walk to them and ask what they'd pay for *right now*, and structure the first offer so some payment lands **before** the build (deposit-backed audit or discovery sprint — the services analogue of a paid waitlist). Wizard-of-Oz applies directly: deliver the first CV-inspection value semi-manually before automating the pipeline. Composes with the phone-mining/ask-past-the-person move below — that supplies the prospect list, this supplies what to test on it.
|
||
- **Two cheap self-tests from the vault's newest voice (added 2026-07-26).** [[2026-07-26-how-to-build-a-billion-dollar-company-2027]] ([[oskar-hartmann]]) supplies founder-facing checks that apply to him directly: (1) the **SOM question** — not "CV/embedded services" but the narrow segment he can win *now* (his industrial-equipment/machine-builder direction is exactly the "AI agent for HVAC contractors" shape Hartmann endorses); (2) the **pricing-power probe** — raise the price on the next quote and watch, the cheapest PMF test available ([[pricing-from-value]]). The same source also seconds his implicit funding default: a niched services operation is in the 99% venture doesn't fit, and per [[venture-fit]] that is a sound place to be, not a failure. Caveat for the partnerships route he's been pointed at: Hartmann's warning that partner hopes disappoint (and whale clients stall past a solo operator's runway) argues for *many small partner bets* over one anchor partner — see [[partnerships]].
|
||
- **A rival diagnosis of the same blocker (added 2026-07-20).** [[technical-founder-trap]] proposes that the missing skill for a technical operator is not networking but **explaining what you do** — solving comes free, articulating doesn't. On this reading his LinkedIn/articles effort failed less because the channel is wrong than because the message wasn't yet legible to a non-engineer. The two diagnoses are testable against each other and imply different fixes (show up in rooms vs. get reps at explaining); the vault has no evidence to choose. Note also that a referral only travels if the referrer can *repeat* what you do in a sentence — which makes articulation upstream of the relational engine too, not an alternative to it ([[referrals]]).
|
||
- **Skeptical of open source** as mostly marketing ("gambling for questionable results").
|
||
|
||
Notable work named in the interview: **insin** (an apt-get-style edge/device updater) and **Keller**.
|
||
|
||
## Evidence
|
||
|
||
- Role, background, harness demo, and the networking problem — [[2026-07-06-sebastian-interview-ai-and-software-engineering]]
|
||
- "Fundamentally I'm Eugene — I'm not a programmer. But I need to work on that." (agreeing with the identity-decoupling point)
|
||
|
||
## Related Pages
|
||
|
||
- [[sebastian]] — his interlocutor
|
||
- [[future-of-engineering-work]] — the BYO-harness thesis
|
||
- [[relationships-as-moat]] — the connections problem he's trying to solve
|
||
- [[client-acquisition-channels]] — his LinkedIn/content investment vs Sebastian's in-person rule
|
||
- [[technical-founder-trap]] — the rival diagnosis of that same blocker
|
||
- [[marketing-system]] — whether his acquisition should be a machine he can turn up, and when
|
||
- [[partnerships]] — the lever that may fit him best: networking with a partner-archetype criterion instead of publishing
|
||
- [[sell-before-build]] — the validation checklist runnable against his offer before any build
|
||
- [[2026-07-26-eugene-90-day-plan]] — the dated plan assembled from all of the above
|
||
- [[2026-07-17-best-method-first-client]] — its undated predecessor (method, not schedule)
|
||
|
||
## Contradictions / Uncertainty
|
||
|
||
- **Likely the vault owner.** `Status: tentative`. Internal evidence: the interview names his project **insin** (apt-get-style edge updater) and his custom Claude Code harness; the owner's environment includes `insin` tooling and this vault is maintained through a Claude Code harness. Strong but circumstantial — recorded as inference, not asserted as fact.
|
||
- Surname not given; "Eugene" is the only handle available.
|
||
|
||
## Next Questions
|
||
|
||
- ✅ **Resolved 2026-07-26** — the dated 90-day plan now exists: [[2026-07-26-eugene-90-day-plan]] (one lever / three venues, warm mining → recurring industry room → partner archetype; paid diagnostic as the smallest paid test; Day-90 gate on inputs). It supersedes the undated method answer in [[2026-07-17-best-method-first-client]] without replacing it. **The plan's own weakest joints are now his open questions:** which specific association/trade fair meets often enough for the 3×/6× mechanic, and whether an industrial buyer pays for a diagnostic at all.
|
||
- **The two rival diagnoses of his blocker are now testable, cheaply.** The 90-day plan's Week 2 phone mine doubles as the discriminator: contacts who engage but can't restate what he does ⇒ [[technical-founder-trap]]; contacts who restate it but know nobody ⇒ [[relationships-as-moat]]. Vault synthesis, no source proposes it — but it costs an afternoon and the vault has had no way to choose between the two diagnoses since 2026-07-20.
|
||
- Did the in-person-networking advice change his channel allocation? (No follow-up source yet.)
|