Screenshots
The personal demo
in action.
From the instance built around my profile — ranked queue, fit scoring, and company-level reasoning. UI still shows the internal build label; the product name is Opportunity OS.
AI · Discovery · Python
A system that finds opportunities aligned to who you are — not just what's open. Ranked discovery, transparent fit scoring, and ready-made email and LinkedIn drafts to apply. The first instance was built around my profile as the live demo.
The idea
Most career tools start with vacancies. Opportunity OS starts with who you are — your themes, capabilities, geographies, and the kind of work that actually fits — then surfaces environments where you could create disproportionate value.
The goal isn't matching keywords on a résumé. It's finding founders, companies, and emerging roles where a conversation could turn into something real — advisory work, a founder-in-residence path, a strategic partnership — and handing you drafts ready to send when you're ready to move.
I built the first instance for myself: my multipotentialite profile, my priority themes, twelve curated companies across six countries, and a pipeline from discovery through outreach. The case study shows what the system does — demonstrated honestly on my data, not a generic product pitch.
How it works
Three layers work together: a profile that defines what "fit" means, a discovery and scoring engine that ranks opportunities, and an action pipeline that generates apply-ready outreach.
01 · Profile
Define fit
Capabilities, priority themes, geographies, and role patterns — the source of truth for what aligned work looks like.
02 · Discover
Score & rank
Curated and agent-discovered companies get transparent scores — theme, capability leverage, role emergence, geography — plus why you fit and problems you could solve.
03 · Apply
Draft & send
Email and LinkedIn drafts, contact targeting, and a focus flow: overview → edit draft → send. Starting points — you edit before anything goes out.
What it found for me
Examples from the personal build — scored companies and generated outreach. Suggested roles are strategic hypotheses; they don't imply the company is hiring.
Graphy · Creator Economy
Why fit: I teach on Graphy — UI friction and creator pain firsthand. Creator + educator + entrepreneur lenses in one place.
Suggested role: Creator Education Founder in Residence
Problems to solve: Educator onboarding from first course to revenue; AI-assisted course creation without losing community depth.
RemotePass · Future of Work
Why fit: Remote work + complex systems — product, ops, and employer journey solved together. Multipotentialite as asset, not liability.
Suggested role: Future of Work Entrepreneur in Residence
Problems to solve: Fragmented employer/worker journeys; compliance + payroll + experience made smarter without losing trust.
Kyan Health · Wellness & Mental Wellbeing
Suggested role: Wellbeing & AI Product Strategy Advisor — targeted to CEO Vlad Gheorghiu with a concrete advisory angle.
Email draft (excerpt):
Under the hood
A background discovery agent researches new candidates on a schedule — countries, themes, industries configurable — and drops them into a review queue before anything enters the ranked list.
A learning layer captures likes, neutrals, outcomes, and explicit answers. Rankings adjust by a capped learned signal (+/-10) without replacing the original editorial score — so the system improves without rewriting history.
Contact and email finder automations help target the right person before drafts go out. Company chat uses Claude or OpenAI when configured; otherwise it falls back to grounded snippets from the data layer.
Tech & status
Streamlit dashboard, Python engines, CSV/JSON data layer — no database required for the personal instance. Cursor Cloud automations handle discovery, contact research, and dev feedback loops.
This is a personal build demonstrated on my profile — not a public multi-user product yet. Source is open on GitHub; a hosted walkthrough is available on request. Drafts are starting points; human review before send is assumed.
Screenshots
From the instance built around my profile — ranked queue, fit scoring, and company-level reasoning. UI still shows the internal build label; the product name is Opportunity OS.
Source & walkthrough
The repo shows how Opportunity OS works end to end — profile, scoring, drafts, and automations. Want to walk through the live personal instance? Email or WhatsApp.
Want a portfolio like this? I'm giving away 5 free — email me and I'll follow up with next steps.