💰 Fundraising Dataroom
Everything on Specky, in one place
Thank you for your interest in Specky. This page is designed to give you a clear, structured, and honest view of the product, traction, team, and fundraise — kept concise and factual rather than polished for effect. If anything's missing or you want to go deeper on something specific, reach out directly.
Specky is building the AI-native workspace for product managers — closing the loop from scattered customer signal to a shipped, cited spec to a measured outcome. Building got cheap; knowing what's actually worth building didn't. Most PMs spend roughly 80% of their week on overhead that produces zero product insight. Specky's bet is flipping that ratio.
FAQ
Questions investors ask
How is this different from Claude or ChatGPT connected to my existing docs?
A generic AI assistant still requires you to gather the signals, decide what's relevant, and remember to check back. Specky ingests Slack, Jira, GitHub, and support signals into one Product Graph automatically, synthesizes themes with citations back to the real source, drafts the spec grounded in that evidence, stages the tickets, and after you ship, checks whether it actually moved the metric — closing a loop no assistant bolted onto your existing stack does end to end.
Two US-funded companies just entered this space — why does Specky win?
Modem.dev ($4.4M pre-seed, Accel) auto-triages feedback for engineering teams; Samepage.ai ($4.85M seed, Craft Ventures) pushes monitoring digests to product leaders at orgs running Jira, Gong, and Salesforce. We read that $9.3M as category validation, because both stop where the actual job starts: neither drafts the spec, records the decision, runs the experiment, or checks the outcome. Specky owns that full loop. Structurally, they also can't chase our buyer — Samepage's integration stack presumes a mid-market org with a CPO, and Modem meters pricing per signal ($144+/mo) where Specky is a flat €99 a solo founder can put on a card.
What stops a competitor from quickly copying this?
Individual pieces — RAG over documents, ticket generation — are replicable by any team with API access to an LLM. What compounds is the Product Graph itself: every signal, decision, and outcome recorded over time makes the next recommendation sharper for that specific team. That's a data moat that grows with usage, not a feature you can ship over a weekend.
How do you keep the AI from hallucinating?
Every insight and generated spec cites back to the real source signal — a Slack message, a support ticket, a call transcript, a synced Figma file. Nothing is presented without a verifiable receipt. SpecBench is our own public benchmark scoring exactly this kind of groundedness, open for anyone to check.
Who's the ideal customer?
Solo technical founders and lean AI-native teams (2-10 people) running product without a dedicated PM — the people building in Cursor, Claude Code, Bolt, Lovable, and v0, who can't afford to guess wrong about what to build next.
Isn't the PM tools market saturated already?
The category is estimated north of $7.5B and every AI-native software team now needs the judgment function whether or not they have a dedicated PM — that's the actual expansion, not a fight over existing Productboard/Jira seats.
What does the cap table look like?
Fathy Shalaby is sole founder, 100% owner pre-raise. Full cap table detail, scenarios, and the financial model are shared directly once we're in a real conversation — see Financial model below.
Let's talk
Fathy Shalaby, founder — Vienna. Happy to go deeper on anything above.