diff --git a/projects/business/memory/proposals-v2.md b/projects/business/memory/proposals-v2.md
new file mode 100644
index 0000000..b9ac05e
--- /dev/null
+++ b/projects/business/memory/proposals-v2.md
@@ -0,0 +1,81 @@
+# Concrete API Product Proposals — v2 (2026-02-14)
+
+## Proposal 1: **Webhook Relay & Transform API** ("hookpipe" or similar)
+
+**What:** API that receives webhooks, transforms/filters payloads via user-defined rules (JSONPath, templates, conditions), and forwards to one or more destinations. Think Zapier's webhook step, but as a standalone API.
+
+**Who pays:** Developers integrating services that don't talk to each other natively. SaaS builders connecting third-party webhooks to their apps. DevOps teams routing alerts.
+
+**Competitors:**
+- Zapier/Make — overkill & expensive ($20+/mo) for just webhook routing
+- Pipedream — developer-focused but complex, requires writing code
+- Hookdeck — closest competitor, $25/mo for 100K events
+- Svix — webhook sending (not receiving/transforming)
+
+**Differentiation:** Dead simple. No code, no workflows, no visual builder. Just: receive → transform (JSONPath template) → forward. Config via API/dashboard. Free tier generous enough to hook devs, paid when volume grows.
+
+**Pricing:** Free 1K events/mo → $9/mo 10K → $29/mo 100K → $79/mo 1M
+**Build cost:** €0 (Go service, deploy on fly.io free tier or Hetzner €4/mo)
+**Time to MVP:** 1-2 weeks
+**Revenue potential:** $500-2K/mo at scale (long tail of small users)
+**Risk:** Medium. Hookdeck is well-funded but expensive. Our angle is simplicity + price.
+
+---
+
+## Proposal 2: **Markdown/HTML to PDF API** ("pdfpipe" or similar)
+
+**What:** Send Markdown or HTML+CSS, get back a pixel-perfect PDF. Focused on documents: invoices, reports, receipts, contracts. Not a general PDF toolkit — just rendering.
+
+**Who pays:** SaaS apps generating invoices/reports. Devs who hate wkhtmltopdf. Agencies automating document generation.
+
+**Competitors:**
+- wkhtmltopdf — open source but abandoned, buggy
+- Puppeteer/Playwright — self-hosted headache
+- DocRaptor — $15/mo (125 docs), Prince-based, excellent but expensive
+- PDFShift — $9/mo (250 docs)
+- Gotenberg — open source (Chromium-based), no hosted offering with good DX
+
+**Differentiation:** 
+1. Markdown-first (competitors are HTML-only)
+2. Built-in invoice/report templates (send JSON, get PDF)
+3. Cheaper than DocRaptor, better DX than self-hosting Gotenberg
+4. Template library for common use cases
+
+**Pricing:** Free 50/mo → $7/mo 500 → $19/mo 2K → $49/mo 10K
+**Build cost:** €0-4/mo (Chromium on small VPS or fly.io)
+**Time to MVP:** 1 week
+**Revenue potential:** $300-1.5K/mo
+**Risk:** Low. Well-understood problem, many bad solutions, room for good DX.
+
+---
+
+## Proposal 3: **JSON Schema Validation API + Registry** ("schemacheck" or similar)
+
+**What:** API to validate JSON payloads against schemas, store/version schemas, generate documentation. Webhook validation as a service.
+
+**Who pays:** Teams building APIs who need payload validation. Webhook providers who want to validate outgoing payloads. Companies standardizing data contracts across microservices.
+
+**Competitors:**
+- ajv (library, not a service)
+- JSON Schema Store (free, community, no API)
+- Stoplight/Optic — API design tools (different category)
+
+**Differentiation:** No direct SaaS competitor for "schema validation as a service." It's a gap because most people use libraries. But an API makes sense for: webhook validation, cross-language teams, CI/CD pipelines, no-code tools.
+
+**Pricing:** Free 1K validations/mo → $9/mo 50K → $29/mo 500K
+**Build cost:** €0 (pure Go, no browser needed, tiny resource footprint)
+**Time to MVP:** 3-5 days
+**Revenue potential:** $200-800/mo (niche)
+**Risk:** High. Market might be too small — people just use libraries.
+
+---
+
+## Recommendation: **Proposal 1 (Webhook Relay)** or **Proposal 2 (PDF API)**
+
+**Webhook Relay** has the best market dynamics: clear pain point, expensive competitors, sticky product (once integrated, hard to switch). But harder to get first customers.
+
+**PDF API** is easier to ship and easier to get first users (developers searching "html to pdf api" is a proven search query with volume). Lower ceiling but faster to revenue.
+
+I'd start with **PDF API** — ship in a week, validate demand, then potentially add webhook features later as a "developer tools" brand.
+
+**Awaiting human decision.**
diff --git a/projects/business/memory/research/competitors.md b/projects/business/memory/research/competitors.md
new file mode 100644
index 0000000..5762619
--- /dev/null
+++ b/projects/business/memory/research/competitors.md
@@ -0,0 +1,19 @@
+# Competitor Research — 2026-02-14
+
+## Screenshot APIs (CROWDED)
+| Service | Cheapest Plan | Per 1K images |
+|---------|--------------|---------------|
+| ScreenshotOne | Free 100/mo, then paid | ~$10/1K |
+| Urlbox | $19/mo (2K) Lo-Fi | $9.50/1K |
+| HCTI | $14/mo (1K) | $14/1K |
+| Microlink | Free tier, then paid | ~$10/1K |
+
+## Image Generation from Templates (CROWDED)
+| Service | Cheapest Plan |
+|---------|--------------|
+| Bannerbear | Credit-based, ~$49/mo |
+| Placid | Credit-based, templates |
+
+## Verdict
+Screenshot/OG space is saturated with well-funded competitors.
+Need to find a niche they don't serve well or an adjacent problem.
diff --git a/projects/business/memory/sessions.md b/projects/business/memory/sessions.md
index 1557ba4..4d3e77e 100644
--- a/projects/business/memory/sessions.md
+++ b/projects/business/memory/sessions.md
@@ -9,3 +9,16 @@
 - **Status:** Sent proposals to user, awaiting approval
 - **Next:** Build MVP once direction is chosen
 - **Blocker:** Need human decision on which model to pursue
+
+## Session 2 — 2026-02-14 12:21 UTC (Morning Session 1)
+- Phase 0: Deep market research on specific API products
+- Scraped pricing from: ScreenshotOne, Urlbox, HCTI, Microlink, Bannerbear, Placid, DocRaptor, PDFShift, Hookdeck, JsonLink, OpenGraph.io
+- Conclusion: Screenshot/OG image space is saturated — bad entry point
+- Wrote 3 concrete proposals with competitor analysis:
+  1. **Webhook Relay & Transform API** — receive, transform, forward webhooks. Hookdeck competitor at lower price.
+  2. **Markdown/HTML to PDF API** — markdown-first PDF generation. DocRaptor competitor with better DX.
+  3. **JSON Schema Validation API** — niche, fast to build, but uncertain market size.
+- Recommendation: PDF API (fastest to ship, proven search demand) or Webhook Relay (stickier, better long-term)
+- **Status:** Proposals v2 written, awaiting human decision
+- **Next:** Build MVP once product is chosen
+- **Blocker:** Need human to pick a product direction
diff --git a/projects/business/memory/state.json b/projects/business/memory/state.json
index ba977d1..a329f51 100644
--- a/projects/business/memory/state.json
+++ b/projects/business/memory/state.json
@@ -1,10 +1,10 @@
 {
   "phase": 0,
-  "phaseLabel": "Business Model Discovery — Needs Specific Product",
-  "status": "needs-specific-product",
-  "currentPriority": "Human approved API direction but needs a SPECIFIC product. Not 'an API' — exactly WHAT the API does, WHO pays for it, WHY they'd pay YOU instead of alternatives, and what competitors ALREADY exist. Do real market research. Find a gap. Propose 2-3 concrete API products with competitor analysis, pricing research, and differentiation. Be specific enough to start coding tomorrow.",
-  "humanFeedback": "User said: 'api sounds good, but the question would be what api? that is not a real business model.' — Need to commit to a specific product, not a category.",
-  "blockers": [],
+  "phaseLabel": "Business Model Discovery — Specific Proposals Ready",
+  "status": "awaiting-human-decision",
+  "currentPriority": "Human needs to pick between Webhook Relay API, PDF API, or Schema Validation API. Proposals in memory/proposals-v2.md with competitor analysis, pricing, and differentiation.",
+  "humanFeedback": "User said: 'api sounds good, but the question would be what api? that is not a real business model.'",
+  "blockers": ["Waiting for human to choose a specific product"],
   "startDate": "2026-02-14",
-  "sessionCount": 2
+  "sessionCount": 3
 }