Every "vertical-AI-for-X" wedge that surfaces across the daily editions gets logged here, tagged with market signals, your-fit score, and a status. Closes the loop between "interesting idea in Tuesday's edition" and "Saturday refinement / decision."
Status legend: 💡 noted · 🔍 researching · 🟡 prototyping · 🟢 active project · 🔴 closed (won't pursue) · 🏁 shipped
Fit score (1–5): 5 = your strengths + market wide-open; 1 = poor fit or market closed
Last updated: 2026-05-19
| Status | Wedge | Buyer | Contract size | Anchor competitor(s) | Fit | Notes / Source |
|---|---|---|---|---|---|---|
| 💡 | Claude + multi-jurisdiction 1099/contractor tax-classification | Small accounting firms (5–25 staff) | $750/mo per advisor | None at SMB tier (Intuit too generic) | 5 | Workday Solopreneur Accelerator candidate. 2026-05-19/05 §5 |
| 💡 | Claude + small-firm M&A due-diligence checklists | Boutique investment-bank associates | $1,500/mo per seat | iManage / NetDocuments (enterprise-only) | 5 | Workday Solopreneur Accelerator candidate. 2026-05-19/05 §5 |
| 💡 | Vertical-CX agent for regulated FS / healthcare / legal | Mid-market FS / healthcare / legal | $5–25K/mo | Sierra ($15B) — enterprise only | 4 | Vertical wedge under Sierra's price floor. 2026-05-19/02 §2 |
| 💡 | Agent identity / KYC / auth layer | Agent-platform vendors | $0.001 per request | None — green-field | 4 | Adjacent to Parallel Web Systems. 2026-05-19/02 §3 |
| 💡 | Agent observability / per-agent endpoint analytics | Production agent teams | $200–2000/mo | New Relic / Datadog (none agent-native) | 4 | Adjacent to Parallel Web + TrajAD research |
| 💡 | JADE-style per-claim eval API | Vertical-AI vendors (Claude for X) | $500–5K/mo | Judgment Labs ($32M Lightspeed) | 4 | Build on JADE methodology. 2026-05-19/04 §2 |
| 💡 | Trajectory verifier API (TrajAD-style) | Production agent teams | $0.001–0.01 per check | None published yet | 4 | First-mover advantage; Haiku verifier + Opus main agent ratio. 2026-05-19/04 §3 |
| 💡 | Cost-aware multi-provider model router | Any team running 3+ providers | $0.0001 per routed token | OpenRouter (consumer-first), early-stage commercial | 4 | Doubles as audit-tool. 2026-05-10 and your personal billing audit |
| 💡 | Public MCP server + connector marketplace | Solo devs + small AI teams | $99/mo per connector | None yet (Anthropic owns spec, not marketplace) | 3 | High visibility, low contract size; pure portfolio play |
| 💡 | Fine-tuning-as-a-service for non-LLM modalities (audio / video / time-series) | ML teams without infra | $1–10K/mo per model | Runware (LLM-focused) | 3 | Adjacent to Runware. 2026-05-19/02 §4 |
| 💡 | B2B onboarding / compliance auto-course generator | SMB HR + compliance | $300–1500/mo | Workday Learning (enterprise) | 3 | Adjacent to Oboe. 2026-05-19/02 §5 |
| 💡 | Vertical-PM-agent for mid-market PMs | Mid-market product managers | $50–200/mo per user | Linear AI features (enterprise-built-in) | 3 | Suleyman 4-verticals lane (PM was the most under-built). 2026-05-18/05 §5 |
| 💡 | AI-tooling for FDEs (Cursor-for-FDE workflows) | OpenAI / Anthropic / Palantir FDEs | $50–200/mo per FDE | None | 3 | Niche but high-fit-with-trends. 2026-05-19/05 §2 |
| 💡 | AI-skills wage-premium calculator + career-tracker | CS students + early-career engs | Free + premium $20/mo | Levels.fyi (not AI-skill-specific) | 2 | Consumer; harder to monetize |
| 💡 | On-device inference orchestration (Apple Extensions / Aluminium OS) | Mobile app devs | Per-device licensing | None | 2 | Requires deep mobile + on-device ML expertise |
| 💡 | AI-drug-discovery startup | Big Pharma | $multi-million pilots | Isomorphic ($2.6B) + Recursion + Insilico | 1 | Too capital-intensive without bio co-founder. Apply to Isomorphic roles instead. 2026-05-19/02 §1 |
| 💡 | Generic customer-experience agent | Mid-market enterprise | $5–50K/mo | Sierra ($15B) + Decagon + Cognigy + Ada | 1 | Category closed |
| 💡 | Generic FDE-as-a-service | Mid-market enterprise | $10–100K/mo | OpenAI Deployment Co + Anthropic Solutions + Palantir + Big-4 | 1 | Category closed; only vertical-FDE specialists remain open |
| 💡 | Generic agent-search infrastructure | Agent-platform vendors | TBD | Parallel Web ($230M) + 1-2 stealth | 1 | Anchored |
| Wedge | What needs to happen first |
|---|---|
| EU-sovereignty AI vertical | Mistral capital event > $5B or formal EU CAISI mirror launches |
| Defense AI tooling | Anduril / Helsing publish public-tooling spec |
| Climate / carbon AI vertical | First $200M+ four-corner round in carbon space |
By Saturday May 23, pick one of the top-3-fit wedges above and produce a 1-pager covering:
- Problem statement (3 sentences)
- Buyer + budget (1 sentence)
- Wedge product (3 sentences)
- First 5 customers to call (list)
- What you would build in 4 weeks to validate
The 1-pager is the input to the Workday Solopreneurship Accelerator application + becomes a permanent artifact in your interview portfolio regardless of outcome.
- Update cadence: every Tuesday after the daily edition + every Saturday at refinement time
- Demotion rule: if a wedge sits 💡 for 30 days without being researched, demote to a
parking-lot/section or close - Promotion rule: any wedge that hits 🟡 prototyping for 14+ days should either advance to 🟢 with a named first customer, or drop back to 💡 with a one-line "why I'm pausing"