Jede Idee musste einen zahlenden Käufer, Belege aus der Diskussion und eine echte Lücke vorweisen und dann ein Red Team überstehen, das sie zu Fall bringen wollte.
Gesichtete Storys
25
Gründlich gelesen
6
Vorgeschlagene Ideen
12
Übrig geblieben
2/12
2 Ideen übrig
| # | Wertung | Idee | Typ | Kategorie | Schmerz | Zahlen | Timing | Lücke | Graben | Bauen |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 6.6 | Skalierbare Plattform für mündliche Prüfungen an HochschulenSoftware und Workflows, mit denen Fachbereiche kurze, strukturierte mündliche Prüfungen (Vivas) in großem Umfang durchführen können – das Prüfungsformat, das KI nicht fälschen kann. | Venture | Bildung | 7 | 6 | 8 | 7 | 6 | 6 |
| 2 | 6.0 | Starter-Kit-Dienst für On-Prem-Betrieb offener ModelleProductisiertes Beratungs- und Tooling-Angebot, das mittelgroße Unternehmen in wenigen Wochen von null zu einem funktionierenden On-Prem-Deployment offener LLMs bringt. | Indie | KI-Infrastruktur | 6 | 6 | 7 | 6 | 4 | 7 |
Verworfen · 10 · unter der 6.0-Linie
| 5.9 | Inline PR annotations instead of review reports | Scored 5.9, under the 6.0 bar. |
| 5.9 | Pre-deploy blast radius and on-call router | The supporting evidence is essentially one engineer who hand-built an internal tool, and stitching diff-reach analysis to PagerDuty ownership is a hard PagerDuty-Startup feature rather than a defensible venture company. |
| 5.9 | LLM provider policy and privacy posture monitor | Scored 5.9, under the 6.0 bar. |
| 5.3 | Continuous model swap advisor for LLM workloads | This is the same product as the token-spend router candidate (re-evaluate models against your prompts, route to cheapest acceptable), and OpenRouter plus Artificial Analysis already supply most of it free. |
| 5.2 | Contradiction-aware memory store for LLM agents | Contradiction detection and decay are roadmap features Mem0 and Zep can add within quarters, and agent builders mostly stay on provider-native memory unless the pain forces migration, making this a feature underlying platforms will absorb. |
5 weitere verworfene zeigen
| 5.2 | In-person exam workflow for handwritten STEM assessment | Gradescope and Crowdmark already do scan-and-rubric grading of handwritten STEM exams at exactly this price point, leaving only marginal workflow tailwinds and no durable gap an incumbent can't close in one release. |
| 5.0 | Token-spend router across frontier and open models | Cheapest-model-acceptable routing is the flagship feature set of OpenRouter, LiteLLM, and Portkey, and the 'quality-eval' differentiator is a feature Martian and the eval tools already bolt on — this is not a company. |
| 4.8 | Benchmark suite for context management strategies | The demand signal is one commenter wishing for this benchmark, the natural outcome is a published comparative report (freely consumed) or a Braintrust eval template, and a $5k one-time report to a handful of platform teams is not a company. |
| 4.7 | Offline skin lesion triage while awaiting dermatology | Too weak on moat (3/10). |
| 4.2 | Turn-key narrow classifier builder for image triage | AutoML and Roboflow already train narrow classifiers from small labeled sets, and 'here's proof it beats GPT on your task' is a sales demo, not a product companies will pay $800/mo for. |