About
I have been a trainer since 2015, with more than 250 cohorts and over 3,000 people trained, and self-taught in everything else. I have trained in multinational corporate and industrial settings, both technical teams and people with no technical background. Along the way I led teams of up to 100 people.
Over the past five years I went from writing scripts to automate repetitive work to building software that runs in production. I started out of necessity, not by plan: checking tools, experiments, and later larger systems. Today, in my own studio, I build blockchain protocols, on-chain payment infrastructure across several networks, and AI systems, from LLM content moderation to writing assistants.
I build and run projects, but none of them pays my bills yet. Knowing how to build is only half the problem; the other half is adoption. That distance is what I wrote about in Quando a Ideia Não Basta and Sem Filtro, both in Portuguese.
I check everything I build three times over. Before anything reaches production, I put it in front of several AI agents whose job is to break it. A fault only counts as fixed once a test reproduces it. It is the same quality discipline I applied for almost a decade to assessing people, now with agents in place of the assessors and a test in place of the scorecard.
Teaching is what I most like doing; building is what gives me something worth teaching. This hub comes out of that combination: the guides come from real training work, and the tools are built with the same care as the professional projects.
The most complete piece of that combination is already published: 8 full technical courses I built on my own, 76 chapters in total, 4 on Python for test automation and 4 on AI engineering (production AI APIs, RAG, evaluation and observability, agents), with automatic progress and certificates, and lessons where the code actually runs in the browser instead of just being shown. Free, with optional voluntary support. They are written in Portuguese.
I am based in Portugal and available for technical training, enablement and AI engineering roles, in Portugal or abroad, hybrid or remote. The studio is not an exit plan: it has always run alongside a full-time job, and it is the space where I keep my hands in the code, which lets me teach from what I have built rather than only from what I have read. The fastest way to reach me is LinkedIn.
Track record
- Full-Stack Developer / AI Engineer, Moto Spotters (LusoCryptoLabs) (2026). A community platform for motorcyclists in Portugal, live in production (Next.js, PostgreSQL), including an LLM content moderation system covering more than 25 content types, a ride passport with stamps verified by the organiser, motorcycle club management, and a peer-to-peer marketplace.
- Full-Stack Developer / AI Engineer, Scryve (LusoCryptoLabs) (2025). A long-form publishing platform, live in production, including an AI writing assistant (embedding-based plagiarism detection) and a custom micropayments system.
- Technical founder, LusoCryptoLabs (2023 to present). The studio where I keep my hands in the code, and under which I built Moto Spotters and Scryve. I design blockchain protocols and payment infrastructure across several networks (Cardano, Nervos CKB, Bitcoin, EVM). I built an AI-assisted security auditing tool, and I apply a multi-agent adversarial audit process to smart contract systems on testnet: finding and fixing vulnerabilities before anything moves towards production, each fix validated with a test that reproduces the problem. I also build products under this studio, such as DoutorWeb, a web design agency for small businesses.
- Technical Trainer, Critical Manufacturing (2022 to 2026). Technical training for industrial manufacturing teams: around 23 training runs a year, cohorts of 20 to 30 people, close to 90 runs and more than 1,800 trainees over four years. A blended model: e-learning, then Q&A sessions with up to 50 trainees for the common questions, then hands-on practice split by track with a dedicated trainer for each. I was the trainer for the Modeling track, certified as an Associate Modeler. I also built Python automation tools to cut manual error out of training administration: my first meaningful exposure to applied programming.
- Trainer and Quality Team Leader, Concentrix (2015 to 2022). Induction and product training for new hires, a new cohort every two weeks with 10 to 15 people: around 170 cohorts and more than 1,600 employees trained over seven years, assessing each person's readiness before they went live. After each cohort I stayed assigned to it for four weeks, already in production and often with the next cohort already in the room, closing the loop between training and real results on the floor: anyone showing a serious gap was pulled, coached, and only then returned to their duties. Later I also took on Quality Team Leader: operations management, with Quality Coaches, Trainers and SMEs reporting to me, across teams of up to 100 people.
- Customer Support Agent, SME and Quality Coach, Concentrix (2013 to 2015). I started as a customer support agent. As an SME, I listened to colleagues' calls, assessed them, gave feedback and reported to the quality coach. As a Quality Coach, I was responsible for grading contacts against four indicators (QA score, critical error rate, compliance, first-contact resolution): root-cause analysis over the error categories, calibrations with the client, and one-on-one coaching on the results.
Recommendation
Excellent quality coach and process trainer. Great communication and motivational skills. The kind of work college that I miss in every office I step in. Go Telmo!
Benjamin Krüger, a colleague at Concentrix, public recommendation on LinkedIn (2021). Quoted exactly as written, uncorrected.
Skills
- Training and leadership (since 2015): facilitation, instructional design, coaching and feedback, team leadership, quality assurance, process improvement.
- Applied AI (5 years): LLM integration, prompt engineering, LLM production systems, RAG and embeddings, AI agents and function calling, evaluation and observability of AI systems, AI adversarial auditing.
- Product engineering (5 years): TypeScript/JavaScript, Next.js/React/Node.js, Python, FastAPI, PostgreSQL/Prisma, Docker, AWS, Cloudflare.
- Blockchain and payments (5 years): Rust, blockchain (Nervos CKB, EVM, Bitcoin, Cardano), on-chain payment infrastructure, smart contract adversarial auditing.
Education and certifications
- Six Sigma Black Belt.
- Associate Modeler, Critical Manufacturing.
- Professional Certificate in Information Technology, FCT.
- Computer Science and Informatics, EIT, 2007.
- Managing Team Conflict; Managing Cross-Functional Teams.
- Certified Trainer Qualification (CCP, Portugal), in progress.
Languages
Portuguese (native), English (C2), French (B2), Spanish (B2), Italian (B1).