Product Engineer · AI-Assisted Product Development · Product Operations

I build digital products from real problems.

I define what needs to be built, structure product and system, orchestrate AI-assisted implementation, test, validate and iterate. I add 13 years of operational judgement built under pressure.

View focus

Ana Victoria
Ana VictoriaProduct Engineer — Alicante, Spain
Product · journeys · decisions
Build → Validate, AI-assisted
18years of professional experience
13years of operational judgement
177automated tests + 7 end-to-end suites (NADIE)
4independent digital systems built

13 years of operational judgement included.

Selected product work

Products built end-to-end from real problems.

AQUO is the flagship product: problem, product concept, journeys, AI-assisted implementation, testing and go-to-market. NADIE shows how I approach risk, evidence and responsible AI. STUDY_OS shows the same discipline applied to governed architecture and testing. Business LAB shows productisation from idea to live operation.

Flagship product · 01

AQUO — daily wellbeing, finally with context

A digital wellbeing product that brings daily signals together — water, sleep, energy, cycle, nutrition, mood and routines — so users can understand patterns instead of judging isolated days.

Digital productProduct operationsAI-assisted workflowsUser context
✓Defined functional structure, modules, user journeys, documentation and product decisions from zero.
✓Used generative AI and Make to accelerate documentation, content operations and repeatable workflows.

THE CASE

A deliberate safety-layer test surfaced five silent failures and led to five mandatory release checks.

Five silent failures found · five permanent release checks

Product → Brand → Positioning → Go-to-market. Beyond the product itself, I built a coherent brand system around AQUO — positioning, visual direction and go-to-market materials including a branding deck and a pitch deck. Implementation details available in conversation.

Closed beta, pre-revenue: billing not activated.

Visit AQUO →
Daily signalsSleep · Water · Energy · Cycle

AQUO

Today is not a score. It is context.

EnergySteady
Sleep6h 42m
Water1.8 L
CycleDay 18
Context engineSignals → patterns → useful reflection
Evidence
112DB tables
50+modules
5mandatory release checks
Structural evidence from the product's functional architecture, with per-user data isolation.
Responsible AI product · 02

NADIE — Invisible Exclusion Risk Analyst

A local-first product that analyses digital journeys to detect where people could be confused, blocked, excluded or unable to recover — before the organisation notices.

Local-firstResponsible AIEvidence system
✓Designed an original risk taxonomy, evidence schema and abstention protocol with mandatory human review.
✓Built a working pipeline with deterministic controls: 177 automated tests passing and 7 end-to-end suites.
✓Human review is mandatory before publication.

THE CASE

A release control blocks publication if the product claims anything it cannot demonstrate — the system is not allowed to assert a finding without evidence behind it.

177 automated tests passing + 7 end-to-end suites

Technical beta: no real pilots yet, synthetic scenarios only.

Visit NADIE →
EvidenceObserved element, step or requirement
RiskConfusion, error, abandonment or loss
RecoveryCan the person continue without help?
CorrectionConcrete, proportional and verifiable
NADIEevidence → risk → human review
Analysis pipeline
01Evidence
02Signal
03Finding
04Human decision
05Report
The machine proposes; a person decides. Human review is mandatory before publication.
Product Engineering system · 03

STUDY_OS — Adaptive Study Operating System

System built from zero to plan, execute and adapt study with continuity, traceability and recovery when a session fails.

Supabase/PostgreSQLGoverned architectureRLSGitHub / CI
✓Governed architecture with separate STAGING and PRODUCTION environments and versioned migrations.
✓Supabase/PostgreSQL with RLS, protected GitHub main and CI with 9 checks in the verified technical foundation.
✓Separate integration, RLS and E2E suites; AI accelerates implementation while verifiable controls govern the result.

THE CASE

Later integration introduced a Planner and Learning Engine with a HOY / LEARN / CHECK / FIN session flow.

696/696 integration tests passing · verified Phase 4B checkpoint

Principles: evidence before assumptions, adaptation without punishment, Mastery ≠ Exam Readiness.

1SpecifyProduct principles, requirements, architecture and validation criteria.DEFINE
2GovernSTAGING/PRODUCTION separated, versioned migrations, protected main.ARCHITECTURE
3Build + AIPlanner + Learning Engine, AI-assisted implementation.AI
4ValidateIntegration, RLS and E2E suites; CI with 9 checks.QA
Evidence
9CI checks
696/696Phase 4B integration tests
RLSRow-level security enforced
Structural evidence from the product's governed architecture, with staging and production kept separate.
Product system · 03

Business LAB — from idea to operation

A practical system for capturing requests, structuring information, clarifying priorities and producing reusable documentation with AI and no-code automation.

OperationsGenerative AIMakeDocumentation
✓Converted fragmented inputs into structured briefs, decisions and next actions.
✓Designed repeatable intake and documentation workflows using Make and AI.

THE CASE

A pre-launch audit found a missing payment route and resulted in an eight-point corrective specification.

Eight-point corrective specification · live and operational

Live and operational · no first sale yet.

Visit Business LAB →
1CaptureForms, messages, raw requestsINPUT
2StructureAI-assisted synthesis and classificationAI
3DecideHuman review, priority and ownershipHUMAN
4ExecuteDocumented next actions and follow-upOUTPUT
Sales funnel
01Visit
02Qualification
03Payment
04Delivery
05Follow-up
This is where the leak was: a payment route was missing before launch.
01

Product thinking

Start from the human need, map the journey, reduce friction and turn feedback into decisions.

02

Operational design

Create structure, ownership, documentation and follow-up so good ideas become reliable execution.

03

Applied AI

Use AI where it improves speed and clarity — never where it replaces evidence, context or accountability.

Builder's toolkit

SYSTEMS I BUILD FOR MYSELF

When a task starts repeating, I try to stop solving it manually and turn it into a system. This is secondary evidence of how I behave as a builder — not a catalogue of primary projects.

CONTENT INTELLIGENCE

Instagram Performance Diagnostic

Analytics → diagnosis → next-post recommendation.

CONTENT SYSTEM

Brand-Aligned Prompt Engine

Brand rules → content objective → structured prompt.

FINANCE OPS

Automated Financial Control

Personal · AQUO · Business LAB.

CONTENT PRODUCTION

Reel Production System

Concept → script → caption → production → result.

PRODUCT DISCOVERY

Living Empathy Map

Signals → user understanding → product decisions.

PRODUCT STRATEGY

Living AQUO Canvas

Problem · audience · proposition · assumptions · evolution.

GO-TO-MARKET

Business Rebrand System

Local presence · web · social · creative growth.

Implementation details available in conversation.

Operational judgement

The judgement came before the AI.

For 13 years I worked in urban mobility and service operations. I coordinated services, groups and pickups, resolved incidents in real time and reorganised operations when bookings, transport or the original plan failed. That experience is now a product advantage: I know how to prioritise, decide with incomplete information and protect continuity.

✓Coordinated with hotels, private transport companies, roadside assistance and other operators.
✓Organised group pickups and airport transfers, distributing demand across multiple vehicles.
✓Resolved booking errors, breakdowns and last-minute changes with a focus on continuity.
✓Communicated with domestic and international customers, including years of service to English-speaking clients.
How I work

The method comes before the tools.

The same route through a wellbeing product, a risk analyser and a service line. Seven steps, each one leaving a documented trail.

Seven-step working method with a continuous improvement loop Problem Understand Specify Build + AI Validate Release Improve
  1. ProblemIdentify the real problem and the friction behind it.
  2. UnderstandUnderstand users, context, constraints and evidence.
  3. SpecifyTurn the problem into structure, requirements, journeys and decisions.
  4. Build + AIUse AI as an execution layer while keeping human judgement and accountability.
  5. ValidateTest real behaviour and look deliberately for failure.
  6. ReleaseRelease only when the product can support what it claims.
  7. ImproveTurn relevant failures into permanent controls or improvements.

A relevant failure does not end with a fix: it should leave behind a test, control, specification, documentation or improvement that reduces recurrence.

Capabilities

Where product, execution and operational judgement meet.

I add the most value at the intersection of product thinking, AI-assisted build and validated execution.

01

Product Engineering

Turn needs and problems into functional products and systems.

02

Product Thinking & Specifications

Problem framing, structure, journeys, requirements, documentation and decisions.

03

AI-Assisted Development

Use AI as an execution layer while keeping human judgement and accountability.

04

Testing & Production Validation

Test real behaviour, detect failures and create controls against recurrence.

05

Product Operations

Documentation, feedback, prioritisation, coordination and continuous improvement.

06

Implementation & Automation

Translate needs into workable flows and systems.

MakeGenerative AIChatGPTClaudeSupabaseNotionGoogle WorkspaceForms & SheetsWebhooksNo-code automationProcess documentationGA4

Operational judgement · Documentation · Evidence-led decisions

“My path into product is not traditional. The thread connecting it is simple: I understand the problem, build the system and take responsibility for the outcome.”

My path into product is not traditional. In 2026 I built four end-to-end digital systems across different domains and worked across product definition, functional structure, journeys, documentation, AI-assisted implementation, testing and validation.

Before that, I spent 13 years solving real problems under pressure in urban mobility and service operations. The intersection of product, execution and operational judgement is where I add the most value.

Evidence before confidenceDo not turn inference into fact. Contextualise metrics and actual status.
OwnershipAccountability for the outcome: define, build, verify and iterate.
SystemsFailures should improve the system — solve and leave a mechanism against recurrence.

Looking for someone who can turn ambiguous problems into reliable product and execution?

Open to opportunities across Product Engineering, AI Product, Product Operations, Implementation and selected operations leadership roles.