vitorsoares.co
Building a personal authority platform with strategy, AI agents and continuous iteration.
Period
2026 – present
Role
Strategy, product and content lead
Context
Live experiment, updated as evidence emerges
Approach
Human-led, AI-assisted
Languages
EN + PT
Status
Live · In progress
Why this site exists
Three things drove the decision to build this.
Much of what I know about product, growth, fundraising and commercial decisions came from founders who shared what they were learning. The rest came from making those decisions myself, often without a clear map. This site is a way to contribute to that same space.
A fundamental transition is underway with AI, and most practical conversation about it happens inside a narrow technical bubble. I work at the intersection of product, growth and AI adoption. I want to contribute to a more honest, grounded discussion — especially for founders and professionals navigating this outside major tech centres.
There is also a longer-term interest in what a more productive and resilient European and Portuguese economy looks like. I cannot contribute institutionally, but I can through teaching, mentoring and sharing experience. This site gives that work a durable home and a wider reach.
The positioning problem
The starting point was a genuine problem: how do you make a non-linear trajectory legible without flattening it?
The path was not linear or perfectly planned, but each chapter changed how I approached the next. A career spanning economics, finance, auditing, consulting, startups, SaaS, fundraising, product, growth, AI products, mentorship and teaching in higher education cannot be reduced to a single headline without losing something essential.
The site needed to serve distinct audiences:
- founders seeking mentorship
- startups considering advisory
- event organisers and conference programmers
- universities and executive education programmes
- podcast creators and journalists
- recruiters and executive teams evaluating Vítor's trajectory
- readers interested in AI, product, startups, growth and Europe
I am selectively available for mentorship, startup advisory, speaking and teaching. I am not currently available for fractional leadership, ongoing consulting retainers or hands-on execution projects. That distinction needed to be embedded in the site's architecture, not just stated on a page.
Foundational principles
Early in the project, and refined through repeated iterations, a set of principles emerged to guide subsequent decisions.
Authority platform, not a CV
Represent a body of work and a point of view, not a list of credentials.
Narrative hub, not isolated pages
Every section contributes to a coherent story. No page feels orphaned from the whole.
Selective availability
Clear paths for mentorship, advisory, speaking and teaching — nothing else implied.
Operator voice, not agency voice
First-person, direct, not mediated through a company or brand.
Bilingual by design
English and Portuguese as parallel first-class expressions, not translations of each other.
Claims connected to evidence
Every capability grounded in a case study, a named engagement or a verifiable fact.
Contextual next steps
Logical paths forward for each audience — not a conversion funnel.
Credibility before commercial impact
If a claim cannot be substantiated, it does not appear.
How the project was structured
The case study is deliberately recursive: the platform documents the strategy, decisions and architecture used to build the platform itself.
A working strategic brief was created early and progressively developed into a durable project context: goals, audiences, positioning guardrails, factual history, availability constraints and editorial principles. That context was also incorporated into the Replit project setup, so the Replit Agent could evaluate tasks against the broader purpose of the site rather than treating every instruction as an isolated prompt.
Implementation previews regularly exposed inconsistencies — copy that contradicted a stated principle, a framing that needed revisiting, a claim that required softening. Each review fed back into the content and the strategic context. The process moved continuously between strategy, implementation, review and correction.
Vítor
Source of truth & strategic decisions
ChatGPT
Sparring, challenge & contextual continuity
Replit Agent
Implementation & technical validation
Vítor
Preview, factual review & publication decision
Live site
Evidence as input for the next iteration
From narrative to technical architecture
Three technical decisions followed directly from the strategic ones.
A bilingual architecture required separate canonical URLs (/en/ and /pt/ prefixes) and content written natively in both languages. A single codebase manages both through language-aware routing, keeping versions structurally consistent without separate deployments.
Routes are prerendered to static HTML at build time — not served by a runtime server. Many crawlers and AI agents do not reliably process client-side JavaScript; prerendering ensures they receive complete content. Structured data mirrors the narrative logic: Person and WebSite entities declared globally, each page declaring its type and linking to the WebSite entity via isPartOf.
Decisions and trade-offs
Six product and positioning decisions shaped the site before any technical choices were made.
Breadth without fragmentation
The site covers product, growth, AI, startup building, mentorship, speaking and teaching. Representing that breadth without appearing to be a generic consultant was a central challenge.
Selective availability
Mentorship, startup advisory, speaking and teaching have clear paths on the site. Fractional leadership, ongoing retainers and hands-on execution projects do not.
Capabilities without making them the identity
AEO, SEO, entity schema and landing pages are demonstrated work. They appear here as selected capabilities — not as the central professional identity of the site.
Evidence hierarchy
Products and ventures Vítor directly led carry more weight than smaller execution projects. Reviews and media appear where they support a specific claim, not as generic social proof.
Accuracy over amplification
Several iterations involved reducing or correcting claims that made roles, events or results sound broader than the underlying evidence supported.
AI speed versus human accountability
AI shortened the loop between strategy, implementation, review and publication. Vítor remained responsible for factual accuracy, prioritisation, trade-offs and every publication decision.
Technical trade-offs
Different page types for different semantic roles
About uses ProfilePage (a profile of a person). Non-editorial authority pages use WebPage. Editorial content uses Article. Index pages use CollectionPage where appropriate. Using the wrong type dilutes the semantic signal.
Generated llms.txt, not a static file
The machine-readable summary is written by the build script, keeping it in sync with the site's actual content without a manual update step.
One codebase, two language routes
Both languages live in the same codebase, updated simultaneously. Two separate deployments would create the opposite problem: content and architecture drifting apart over time.
Current state and evidence
The site is live at vitorsoares.co. Baseline content and design have stabilised. Analytics are active — PostHog tracks page views across all routes. Discoverability has been a deliberate focus: prerendered HTML, canonical and hreflang coverage, structured data, sitemap, robots.txt and llms.txt are all in place.
Built and observable
- Bilingual site architecture (English and Portuguese)
- Narrative homepage and About page
- Selected work, case studies and supporting pillar pages
- Prerendered HTML for all public routes
- Canonical and hreflang implementation across all route pairs
- Structured data on most pages (coverage varies by page type)
- Sitemap, robots.txt and llms.txt
- Analytics and page view tracking active across all routes (PostHog)
Not yet proven
- Sustained organic visibility
- Ranking for relevant non-branded queries
- Recognition or citation by AI systems
- Google entity recognition
- Qualified inbound opportunities attributable to the site
- Which visitor paths generate the most useful engagement
What I am learning
The brief matters more than the tools
AI-assisted output quality was directly proportional to the quality of the foundational context. Precise context — who reads this, what they should feel, what they should do next — produces useful output. Vague context produces generic output. This applies equally to briefs written for human teams.
Structural decisions compound
Building bilingually from the start felt like overhead in week one. Weeks later, both versions update simultaneously with no divergence to manage. Decisions at the foundation either pay off or create debt; they rarely stay neutral.
Accurate self-presentation requires discipline
A common failure mode in personal sites is describing the best version of a trajectory while readers look for evidence. Connecting every claim to a concrete story is a discipline, not a default. The case studies section forces broad claims to become specific and verifiable.
What comes next
This case study will be updated in place as the project evolves.
The editorial territory the site will develop over time covers product strategy and B2B SaaS; practical AI adoption and AI-powered products; startup building and go-to-market; growth and commercial systems; negotiation and founder communication; insights from mentoring and teaching; and Europe and Portugal, approached practically and without partisanship. This is not a publishing schedule — it is the evolving scope of the authority platform.
As evidence accumulates — indexing patterns, organic visibility, citation behaviour, inbound quality — it will be documented here. The goal is to record what actually happens, not to project outcomes in advance.
Interested in how I approach products, AI and founder challenges?
This site documents the same principles I bring to mentorship, startup advisory, speaking and teaching: strategic clarity, practical execution and intellectually honest iteration.