PhD + Postdoc
Bioinformatics · UFMG
Verifiable evidenceRECRUITER FAST TRACK · 60 SECONDS
If you're assessing my profile for a role, start here.
Education, professional scope, evidence and navigable work in one short read. The full lab remains available, but you do not need to cross it before deciding whether a conversation is worthwhile.
AI applied to learningLearning DesignAI FluencyLearning AnalyticsNo-code automationTeaching, research and innovation
Professional positioning
A specialist in AI Applied to Education and Learning Design, learning strategist, AI Fluency facilitator and no-code solution orchestrator, grounded in a scientific Bioinformatics background.
I do not position myself as a software engineer or full-stack developer. Code and prototypes are means to materialise learning systems, evidence and automation.
Documents and contact
11 papers
Peer-reviewed scientific publications
Verifiable evidence260+
Academic citations documented in the portfolio
Verifiable evidence10+ years
Teaching, research and learning
Verifiable evidenceTeaching proof
Teaching AI
A navigable micro-lesson with diagnosis, explanation, practice, feedback and learning evidence.
Open teaching proofThe four flagships to inspect
AI Fluency for Educators
An AI fluency framework with a navigable instrument and explicit criteria.
Inspect AI FluencyIntelligent Learning Journey
Adaptive learning architecture and journey simulation with declared decision logic.
Inspect journeyLearningOps Automation
No-code automation with triggers, states, exceptions and explicit human handoff.
Inspect LearningOpsLearning Analytics Command Center
Analytical work connecting learning, signal, evidence limits and business decisions.
Inspect AnalyticsThis view uses only facts and projects documented in the portfolio. Simulations, hypotheses and limitations remain labelled on their source pages.
EVIDENCE MATRIX · CAPABILITY → PROOF → LIMIT
You do not have to trust the capability list. Follow the evidence.
Each row connects a stated capability to Truth Layer records, a navigable proof, the limit of what can already be concluded, and the best interview test for producing new evidence.
01Capability
Teaching AI and facilitation
Teaching and research experience combined with a navigable teaching route containing a micro-lesson, feedback, evidence and a Facilitation Brief.
Open teaching proofRegistered evidence
Verified evidence2026-07-16
Over 10 years in teaching and research
Verified evidence
4 navigable R&D systems
Current limit
The portfolio demonstrates method and artifacts; it does not publish learning metrics from a specific cohort.
Best validation
Ask for a three-minute explanation to a non-technical audience, then introduce an unexpected question.
02Capability
AI applied to learning
Professional experience and four R&D systems connect AI to teaching, learning design, operations and analytics.
Open AI FluencyRegistered evidence
Verified evidence2026-07-16
Over 10 years in teaching and research
Verified evidence
4 navigable R&D systems
Simulated data
AI Fluency for Educators (prototype)
Current limit
The systems demonstrate architecture and reasoning; prototypes do not prove client implementation outcomes.
Best validation
Provide a real learning problem and inspect whether the solution begins with the decision and evidence rather than the AI tool.
03Capability
Learning design and adaptive journeys
Stated Learning Design experience plus navigable systems that make routing and adaptation criteria observable.
Open adaptive journeyRegistered evidence
Verified evidence2026-07-16
Over 10 years in teaching and research
Verified evidence
4 navigable R&D systems
Simulated data
Learning Journey Inteligente (prototype)
Current limit
The adaptive journey has not yet been validated with a real cohort; prototype metrics are simulated.
Best validation
Give two learner profiles and ask for explicit criteria to keep, remove or reorder parts of the journey.
04Capability
Research, method and scientific communication
PhD, postdoctoral work, scientific publications and academic impact have publicly verifiable sources.
Open science layerRegistered evidence
Verified evidence2026-07-16
PhD in Bioinformatics, UFMG
Verified evidence2026-07-16
Postdoctoral research, UFMG
Verified evidence2026-07-16
11 scientific publications
Verified evidence2026-07-16
Over 260 academic citations
Current limit
Academic production should not automatically be treated as operational impact in a corporate context.
Best validation
Ask him to turn a vague hypothesis into a testable question, required evidence and a decision criterion.
05Capability
Learning Analytics and scientific reasoning
Verifiable scientific training combined with an analytical demonstration that makes hypothesis, signal, limits and decision explicit.
Open analytics proofRegistered evidence
Verified evidence2026-07-16
PhD in Bioinformatics, UFMG
Verified evidence2026-07-16
Over 10 years in teaching and research
Simulated data
Learning Analytics Command Center (prototype)
Current limit
The dashboard uses synthetic data and does not represent an analytical outcome obtained for a real client.
Best validation
Present ambiguous data and ask for a competing hypothesis, causal limit and which additional data would change the decision.
06Capability
No-code automation and LearningOps
Operational reasoning demonstrated through trigger, state, exception, retry, diagnosis and explicit human handoff.
Open automation proofRegistered evidence
Verified evidence2026-07-16
Over 10 years in teaching and research
Simulated data
LearningOps Automation (prototype)
Current limit
LearningOps is a demonstrative prototype; there is no documented production outcome or real client saving.
Best validation
Provide a broken workflow and evaluate the investigation sequence before any tool edit is proposed.
No fit score and no automated seniority inference. VERIFIED confirms the described fact; SIMULATED identifies a prototype or demonstrative data, not a client outcome.
DECISION ENGINE · EVIDENCE BEFORE IMPRESSION
What do you need to validate before deciding whether an interview is worth it?
Choose an evaluation question. The system does not calculate a fit score: it surfaces the most relevant proof, what that proof supports, and what still needs validation in conversation or a practical exercise.
Choose the decision you need to make
Priority proof
The strongest proof is a navigable teaching experience: objective, explanation, practice, feedback and a learning check.
Open teaching proofWhat to inspect
Inspect whether the explanation reduces ambiguity, creates useful practice and leaves learning evidence rather than merely delivering content.
What still needs interview validation
Ask for a live micro-explanation for a non-technical audience and test how the approach changes when a real question appears.
Related evidence
Verified evidence
verified 2026-07-16Over 10 years in teaching and research
More than 10 years of professional experience combining teaching, research and the applied use of AI and data science.
What this evidence does not prove
The number of years is an estimate based on the professional record available publicly. Time spent in each specific area may vary.
Verified evidence
verified 2026-07-16PhD in Bioinformatics, UFMG
PhD in Bioinformatics from the Federal University of Minas Gerais (UFMG), one of the most respected graduate programmes in Brazil.
What this evidence does not prove
The doctorate is recorded in the Lattes CV and can be checked publicly. Specific details of the thesis are not available on this page.
Verified evidence
4 navigable R&D systems
Four research and development systems are implemented as navigable pages in this portfolio: AI Fluency for Educators, Learning Journey Inteligente, LearningOps Automation and Learning Analytics Command Center.
What this evidence does not prove
That the four systems exist and can be navigated is verifiable directly in the portfolio. The internal metrics of each one are simulated for conceptual demonstration. See the Content Truth Ledger of each project.
TRUTH LAYER
The portfolio separates verifiable facts from prototypes and simulated metrics, so demonstrated capability and real-world outcome are not treated as the same thing.
No proprietary hiring score. No invented fit percentage.
6verified evidence records
4prototype / simulated records
INTERVIEW PACK · 15-MINUTE PRACTICAL TEST
Do not trust the portfolio alone. Test the reasoning.
This script turns the application into a small validation session. It does not claim experience the portfolio cannot prove. It makes observable how Carlos explains, diagnoses, decides and recognises limits.
Choose the interview focus
Practical exercise
Teach AI to a non-technical person
Give Carlos an AI concept that often causes confusion, such as hallucination, context or response evaluation. Ask for a 3-minute explanation for a non-technical person, then introduce an unexpected question.
Follow-up questions
- 01What did you choose to simplify, and why?
- 02How do you verify that the person actually understood?
- 03What would change for leaders, teachers or operations staff?