Every idea that went through the panel.

“Gather a bunch of professional trader to form a business of investment”

Dead on arrival, no edge, no track record, and regulation will eat you before investors even take the meeting.

“Vision 5 — “Le Acquisition Machine” Tu construis Nexory pour le vendre, pas pour le garder. La vraie vision n’est pas de faire tourner Nexory pendant 10 ans — c’est de le construire proprement, d’atteindre 50 000–100 000€ ARR, et de le vendre sur MicroAcquire ou Acquire.com pour 3–5x ARR — soit 150 000–500 000€ de liquidités en 18–24 mois.[mktclarity] Puis tu recommences avec un second SaaS, plus vite, avec plus de capital, plus d’expérience. La logique : Les SaaS bootstrappés avec un ARR entre $50k et $500k se vendent en moyenne à 3–4x ARR sur les marketplaces de cession. Un solo founder qui construit et vend un SaaS tous les 2 ans génère plus de capital qu’un fondateur qui garde le même produit pendant 10 ans à faible croissance.[mktclarity] Ce que ça change maintenant : Tu construis Nexory avec une architecture propre, documentée, sans dette technique, avec des métriques claires (MRR, churn, LTV) — pas pour toi, mais pour un acheteur futur. Chaque décision de build se pose avec la question “est-ce que ça rend le produit plus vendable ?””

Clean code is not a company, buyers acquire growth and wedges, not tidy architecture built for a flip.

“a webapp that turns regular selfies to a professional photo and resume ready image”

Free tools already do this, and without a specific buyer or distribution wedge, it's a commodity with no reason to win.

“Elearning for 0 knowledge”

Crowded market, no specific learner or outcome, you have teaching instinct but not a business yet.

“An OpenClaw Implementation agency which makes the private servers with open claw in it for bussiness To automate their work, lead gen and do things faster all with a whatsapp or telegram chatbot”

Good problem, generic agency, no wedge, buyers will default to known automation tools unless you own a painful niche.

“Get your brand cited by ChatGPT, Claude, Gemini…”

Real anxiety, imaginary buyer, you are selling SEO to a search engine that owes you nothing.

“An AI-powered reasoning system designed to help people make better decisions, faster”

Observation, not startup, giant models already sell reasoning and you still have no buyer, use case, or wedge.

“Search engine”

Dead on arrival, search needs a sharp wedge, and you have neither a market nor defensible distribution.

“AI-Powered Product Data Optimizer”

Bundled and undefined, this dies before launch unless you pick a buyer, a metric, and one painful optimization job.

“Un SaaS pour la création d’agents et l’utilisateur connecté des outils lui même”

Good insight, no wedge, crowded market, and no defined buyer means this dies before becoming a product.

“Un SaaS pour la création d’agents et l’utilisateur connecté des outils lui même”

Category rouge, wedge flou, seul un agent builder pensé Europe peut éviter de mourir face aux plateformes établies.

“Un SaaS pour la création d’agents et l’utilisateur connecté des outils lui même”

Crowded market, no buyer, no moat, you are building a commodity before choosing who would pay.

“Un SaaS pour la génération de post linkedin”

Marché saturé, cible floue, distribution bloquée, tu te bats contre du gratuit dans un segment déjà verrouillé.

“A form builder, where you can either make the form available public, or restricted access only. Form fields are connected to QR code items, so you can scan QR-codes and get prefilled fields.”

Neat feature, not a company, until you pick a field workflow where scanning beats typing.

“lead gen for reddit”

Platform risk and no wedge, Reddit can copy you before you find a buyer.

“I have a startup idea where users like businesses can book a digital billboard across the major cities. We can install digital billboards and connect it online so when a buisness books a particular slot in the day, their advertisement will show.”

Real demand, brutal economics, build booking software for existing billboards or get crushed installing your own.

“aI Books search engine”

Crowded market, no buyer, no wedge, AI book search dies unless it solves one painful search job better.

“App that connect with your HomeKit device system and matter hardware connected to detect status of the battery and status of the system. Doo open ecc. everything in a glance”

Useful pain, but Apple already owns the glance, so win only by serving messy mixed-protocol homes first.

“a agentic AI like open claw but manages your pets and send you notifications, record walks, pick good food for your pets”

A feature bundled into a consumer app, not a defensible product—and the core job (pet care reminders) is already solved by free, boring tools that work.

“website like a social platform for people to discover them selvs tru tests quizes, using that knowledge ai would suggest them a book that fits their needs, problems, views etc. Scalable into job posting platform (matching perfect candidate for X job, tru detailed analyze from the tests etc.) , dating platform , everyone will get a own profile and use that profile as replace for Instagram, Facebook, LinkedIN, Tinder, Badoo etc..”

Pick one wedge first, right now you're pitching a personality layer searching for a business.

“créer site web de marketing pour aider les entreprises à leur page de réseaux sociaux”

Broad, crowded, already cheap, without a sharp niche or unfair advantage, this dies before the first customer.

“Get diagnostics on your startup to see the real chance of success”

Real anxiety, no buyer, founders will not pay to quantify doubt before they even have a startup.

“AI Chief of staff”

"AI Chief of Staff" is a positioning problem masquerading as a product — you have a feature category, not a wedge.

“Compliance: The "AI-Hiring Disclosure" Automator The Problem: New 2026 regulations (like the updated CCPA and the "One, Big, Beautiful Bill") now require businesses to provide a "Pre-Use Notice" and "Opt-Out" rights if they use AI to evaluate job applicants or employees. Most HR teams are currently breaking this law because they don't have a system to track these disclosures. The Solution: A simple, reliable "Proof of Disclosure" log. It’s a specialized form/widget that sits on their "Careers" page, logs the user's consent to be evaluated by AI, and generates an audit trail. How to Get Paid: Monthly subscription ($49/month) per company. This is "insurance" pricing—they pay to avoid a $20,000 fine. How to Market: Content marketing targeting "Fractional HR" consultants and small-business law firms. When they tell their clients "You need to disclose AI use," they recommend your link.”

Built on imaginary regulation, underpriced for real legal risk, with a referral channel that will not carry you.

“Daily quotes”

Nice human desire, no wedge, no buyer, and free alternatives already own the habit.