# Humanswith.AI

> An operated workspace for AI-search marketing work. Eighteen agents in one workspace. Five agents own marketing responsibilities buyers meet (visibility, content, publishing, website readiness, visuals); support, control-plane, and ops agents complete the install. 10× article throughput: 100–120/mo with one operator + 18 AI agents vs 10–12/mo from a 6-person freelance team. Compared with a 6-person freelance/content team at 10–12 articles/month; one Marketing Operator + 18 AI agents at 100–120 articles/month. Humans retain strategy, approval, exception, merge, deploy, and publishing authority.

## Core pages

- [Home](https://humanswith.ai/): platform overview — the agentic workspace for marketing teams to win in AI search.
- [Marketing Agents](https://humanswith.ai/platform/marketing-agents/): commercial platform page — the expert-guided workspace for AI Search visibility, approved agent work, ContentOS production, Website Agentic Optimization, and weekly re-measurement.
- [Hermes + AI Visibility Intelligence](https://humanswith.ai/platform/hermes/): observation and decision layers that keep mentions separate from citations and turn comparable evidence into governed opportunities.
- [How it works](https://humanswith.ai/how-it-works/): operating-method hub — the weekly workflow from Hermes audit to closeable gap, source-backed work packet, human approval, publication, and re-measurement. Mirrors: https://humanswith.ai/ru/how-it-works/ and https://humanswith.ai/ar/how-it-works/.
- [Marketing Operator](https://humanswith.ai/marketing-operator/): the accountable human role inside the operational pilot — prioritisation, evidence and approval governance, canonical-first publishing, proof, and transfer to a trained client employee after setup.
- [Distribution guardrails](https://humanswith.ai/distribution-guardrails/): canonical-first publishing policy — owned source first, channel-native adaptations, visible source links, bounded internal cadence, stop conditions, and rescan proof.
- [Workspace ContentOS](https://humanswith.ai/platform/contentos/): AI content operations agent — turns AI visibility gaps into source-backed briefs, drafts, schema, quality reports, Workspace Files artifacts, and approval-ready content packets.
- [Content Package Studio](https://humanswith.ai/platform/content-package-studio/): operator-run packaging step — turns an accepted ContentOS article and receipt into a branded PDF package, schema deliverable, manifest, and verified destination handoff.
- [Content Publisher](https://humanswith.ai/platform/content-publisher/): publishing and distribution agent — turns approved ContentOS packets into canonical routes, metadata, schema handoff, distribution drafts, guarded rebuild proof, and follow-up visibility scan tasks. Mirrors: https://humanswith.ai/ru/platform/content-publisher/ and https://humanswith.ai/ar/platform/content-publisher/.
- [Website Agentic Optimization](https://humanswith.ai/platform/website-agentic/): prepare-observe-verify workflow for project-bound website changes, exact-commit checks, human merge authority, and live behavior proof.
- [Visual Asset Studio](https://humanswith.ai/platform/visual-asset-studio/): placement-specific asset rendering with design-pack, fit, safe-area, provenance, and final-pixel verification.
- [Parsing Stack](https://humanswith.ai/platform/parsing-stack/): supporting evidence product for bounded source retrieval and structured facts; it does not decide or publish.
- [Workspace control plane](https://humanswith.ai/platform/governance/): routing, rule compilation, proof, files, memory, handoff, and token policy with current version and shadow-status disclosure.
- [ContentOS Agent Lite](https://github.com/humanswith-ai/contentos-agent-lite): MIT open-source local skeleton of the ContentOS process — seven gates from business context and source pack to draft quality, editorial uplift, and publish readiness. Runs in Claude Code, Cursor, and Codex with zero required API keys.
- [RU Marketing Agents](https://humanswith.ai/ru/platform/marketing-agents/): Russian commercial page for продвижение в нейросетях, ИИ-агенты-маркетологи, видимость в нейросетях, AEO/GEO, and the operator-controlled production loop.
- [Pricing](https://humanswith.ai/pricing): four platform tiers plus a $3,000/mo operational pilot. The pilot separates the Agentic Workspace from the Marketing Operator and makes the human role transferable after setup. The itemized licence comparison totals $1,769.91/mo using only tiers vendors publish; unpriced positions are not estimated or summed.

## About the platform

Humanswith.AI is one operated workspace — eighteen agents (18 installed) coordinated by a Marketing Operator. Five agents own buyer-facing marketing responsibilities across visibility analysis, content production and quality gates, canonical publishing and distribution, technical website readiness and verification and repeatable visual asset delivery; the rest run evidence, packaging, publishing adapters, and the control plane. Humans retain decision authority:

1. **Hermes + AI Visibility Intelligence** [marketing] — Comparable AI-visibility observations, citation vs mention split, and governed next-action handoffs.
2. **Workspace ContentOS** [marketing] — Source-bounded briefs, drafts, schema handoff, quality receipt, and approval-ready packets.
3. **Content Publisher** [marketing] — Canonical-first publishing, channel adaptations, guarded rebuild, and live proof.
4. **Website Agentic Optimization** [marketing] — Project-bound website change packets with exact-commit checks and production verification.
5. **Visual Asset Studio** [marketing] — Placement-specific visual assets with fit, safe-area, provenance, and pixel proof.
6. **Parsing Stack** [support] — Bounded source retrieval, SERP evidence, and structured facts with uncertainty checks.
7. **Content Package Studio** [support] — Branded client PDF packages, schema deliverables, manifests, and destination proof.
8. **Workspace Orchestrator** [control_plane] — Classifies requests, projects authority, builds the execution plan, and records required proof.
9. **Agent Governance** [control_plane] — Compiles project rule packs and records winning vs shadowed rules (shadow mode until enforced).
10. **Proof Loop** [control_plane] — Runs task-relevant tests, static analysis, visual baselines, and golden gates before done.
11. **Workspace Memory** [control_plane] — Keeps sourced decisions, accepted artifacts, install proofs, and audit facts across sessions.
12. **Workspace Files** [control_plane] — Owns shared artifact paths, checksums, and delivery proof operators can open.
13. **Handoff Continuity** [control_plane] — Records continuation state, proof, blockers, and the next executable action across sessions.
14. **Token Profile Router** [control_plane] — Selects the smallest safe tool profile and keeps heavy tooling opt-in for a stated reason.
15. **Proactive Controller** [control_plane] — Suggests safe preflight and stale-install checks without publish, send, delete, or credential mutation.
16. **Medium Publisher** [ops] — Channel adapter: one canonical URL into one Medium draft/publish job under one server-owned session.
17. **MailOps** [ops] — Turns email into governed workspace packets with draft-only reply capture.
18. **Regula SEO** [ops] — Project-bound SEO truth layer when installed for a tenant; not a separately sold generic product.

The commercial Marketing Agents page explains how these modules run as one workspace: measure the gap, choose the approved work, produce with evidence, publish safely, then re-measure the next week. The Content Publisher module is the canonical-first publishing layer between ContentOS and the proof scan. Methodology source: https://gregshevchenko.com/research/ai-what-next-workspace-agents/.

The How it works page is the route-level operating method: https://humanswith.ai/how-it-works/. It connects the product graph without duplicating module pages. The workflow is: run Hermes scan, choose the closeable gap, produce the work packet through Marketing Agents, approve before publishing, then re-measure weekly. RU and AR mirrors are live for Russian-speaking and Gulf/UAE buyers.

The operational pilot targets approximately 300+ content units over the full 3–4-month run. Its steady-state monthly range is 120–160 units, so the all-in unit calculation is $18.75–$25. The upper figure applies only at the conservative output floor and excludes client review time, paid distribution, third-party author costs.

## Proof

- Humanswith.AI dogfooded on itself: 2 → 1,000+ citations in 3 months. 819 AI mentions, 15.4% visibility share — five to ten times ahead of the field.
- GAC (auto retailer, St. Petersburg): from AI-blind to cited on all nine platforms in six weeks. 9,042 reads. "Now ChatGPT recommends us by name." — Head of Marketing.
- Birdview PSA (B2B SaaS, Canada/USA): ChatGPT share 0.9 → 21.5% in eight weeks. ×23 lift. Now cited alongside Monday.com, Asana, Wrike.

## Founders

- **Gregory Shevchenko** — Co-founder, CEO/CTO. 13+ years performance marketing + 8 years building production systems. Founded Humanswith.AI on 2019-07-15. Codes Workspace Core and Agents himself. Speaker: Forbes, WSJ, Tagline Awards. Personal site: https://gregshevchenko.com/. Same person as the Gregory Shevchenko Person entity on gregshevchenko.com (sameAs).
- **Ekaterina Lebedeva** — Co-founder, CFO, Business Development. 15+ years management consulting (Tech & Telco). Ex-CEO of a tech consulting company. Led a 30+ expert team. $100M+ in M&A. Launched Visa- and Gartner-recognized products.

## Cited across

ChatGPT, Claude, Perplexity, Gemini, Grok, DeepSeek, Kimi, Google AI Overviews, Microsoft Copilot.

## Locations

Dubai (HQ), Tbilisi (engineering). Serves clients globally.

## Contact

g@humanswith.ai for partnership and platform inquiries. team@humanswith.ai for everything else.

## Optional

- [Founder's personal site](https://gregshevchenko.com/): Gregory's research, including the MCP-stack token-economy paper, AEO/GEO methodology articles, and primary AI-search analysis.