An autonomous marketing-and-sales platform, designed and built solo over eleven weeks. It runs the day-to-day work of a five-to-seven-person GTM org — pipeline, brand, sales enablement, demand gen, and the data layer underneath — at a measured cloud cost of $25–$50 per month.
The full valuation memo is embedded above. This is the short version: what it cost, what it does, and why the numbers hold up in front of a hostile auditor.
The build
- 11 weeks, April 4 to June 22, 2026 — solo contractor, sole architect.
- $43,603 in labor (692 hours at $63/hr, across six paydays).
- 1,606 commits and +695K net lines of code across nine production work streams.
- 100+ production deploys on
mainvia a Cloud Build auto-deploy pipeline. - Ongoing run cost of $25–$50/month — roughly $2,400/year — with a budget-alarmed ceiling of $200.
The impact
- $245K weighted Y1 pipeline at industry-benchmark close rates, sourced from 135 ICP-fit companies and 47 qualified contacts now live in the CRM.
- ~70% fewer operator clicks per outbound action — capture, draft, send, and log now happen from a single browser tab.
- 8.5× return on labor against the audit-floor valuation.
- Structural cost arbitrage of $1.23M–$2.09M per year versus the staffed alternative — 500–875× cheaper to operate than the org it replaces.
Valuation, anchored
Every lens cites a published benchmark. The replacement-cost figure follows ASC 805, the US-GAAP standard auditors apply to internally-developed software — no internal-assumption math.
- Audit floor — ~$371K. ASC 805 replacement cost (~$126K) plus the benchmark-rate Y1 pipeline (~$245K). This is the number that survives a hostile auditor.
- Durable value — $9.8M–$16.7M. The annual operating-cost arbitrage capitalized at the mid-market 8× SaaS revenue multiple, independent of any forward-revenue assumption.
- 5-year NPV — $7.14M, at a 99% platform-enabled margin and a 12% discount rate.
The reason the run cost is so low is a deliberate architectural decision, not luck: 96% of LLM calls route to Google's open-weight Gemma models — roughly 1/1000th the cost of frontier calls — with frontier models reserved for the 4% of cases that require them.
Six lanes of impact
The 1,606 commits cluster into six business-impact lanes. Each shipped a production artifact the founding team and sales lead now operate on daily.
01 · Pipeline & Revenue Operations. The demand-capture funnel, end to end — from signal detection through CRM hand-off into a single source of truth in Attio. Five production import pipelines, a hiring-signal crawler spanning JobSpy / Lever / Ashby / Greenhouse / Workday / Taleo / BambooHR, a committee-prediction engine with 35 org-shape priors, and a live Attio dashboard (KPI cards, leaky-holes funnel, by-industry breakdowns, today's queue). Result: 135 ICP-fit companies and $245K of weighted Y1 pipeline.
02 · Brand & Category Authority. A content engine that produces hub-and-spoke long-form pillars on schedule, monitors live performance, and self-remediates. 14 clinical-cluster whitepapers drafted, gated, and published to Sanity; a four-module SEO remediation loop that detects, fixes, and re-scores autonomously; an Ahrefs rank tracker on 60 cluster keywords; and a documented, auditable LLM-citation posture. Result: a self-healing content surface.
03 · Sales Enablement & Operator Productivity. Friction removed from the operator's day. A Chrome extension shipped across five releases (v0.10 → v0.14) in eleven weeks — auto-update infrastructure, LinkedIn-parity capture with dedup hardening, a Reddit auto-browse driver with live-progress UI, and cross-operator coordination — plus a slide editor for investor-pitch and internal-deck workflows and three calibrated voice profiles. ~20 internal users. Result: ~70% reduction in operator clicks per outbound action.
04 · Agentic Orchestration Platform. The platform that lets one operator run the work of three to four — the most compounding long-term asset in the build. An always-on agentic caretaker with a permission charter and lane routing; a claim / decompose / finish work-queue framework with a milestone substrate; a skill-adapter registry where every new import vector or composer routes through a central, CI-enforced gate; and an OAuth-managed tick registry with per-surface kill-switches. Result: 703 commits of platform plumbing that every future capability rides on.
05 · Demand Generation Infrastructure. The ad-deployment and attribution pipeline. A LinkedIn document-format ad pipeline with bulk upload, smoke-test tooling, and Lead Gen Form integration; a GA4 campaign attribution model with enforced UTM conventions; and a PostHog event substrate cataloguing 96 event types. Result: an auditable demand pipeline from impression to click to session to MQL to SQL to won.
06 · Data & Knowledge Substrate. The long-term institutional memory and governance layer. Voice corpora for three personas, doctrine catalogs, a PostHog data-quality audit with a contact-enrichment substrate, an ICP-fit scoring engine with a reproducible audit trail, and a country-subdomain dedup utility that caught 29 ghost company records. Result: knowledge that survives compaction, founder turnover, and pipeline pivots.
Stack
Python · TypeScript · React · Tailwind · Sanity · Cloud Run · Firestore · BigQuery · GA4 · PostHog · Ahrefs · Attio · Lemlist · Apollo · Brevo · LinkedIn Ads · JobSpy · Voyager · Chrome MV3 · Playwright · Gemini · Gemma
Firestore as primary OLTP, BigQuery as the analytics warehouse, Cloud Run auto-deployed on every push to main. All external integrations OAuth-managed and routed through a central substrate-integrity guard.