Daniel Oliveira

Five years, four environments.

Research

Universidade de Aveiro. Deep learning driver monitoring on real hardware, end-to-end data pipelines.

Enterprise

Stats Perform and ITGest. Real-time data at scale, enterprise applications, and my first development team to lead.

Startup

Cyberr, Luxembourg. Built the platform from an empty repository and kept it standing past 100,000 registered users.

My company

Nextis Labs. Founder and CTO, setting the technical direction and shipping AI systems.

Five years in production. Three teams led.The stack changes, the motivation does not.

From electronics to applied AI.

Daniel Oliveira — No signal
Portrait slot, 1:1. Supply a photo and it drops straight in.

Electronics
HardwareEmbedded systems
Robotics
Computer visionDeep learning
Applied AI
AgentsMCP serversHarnesses

Experience

2026 → Now

Founder & CTO

Nextis Labs, Portugal

Production AI for clients: RAG systems, LLM integrations, and custom agents, from prototype to deployment and monitoring.

Details

Set the technical direction and keep the architecture matched to what the business actually needs.
Architect and deploy RAG systems, LLM integrations, and custom agents into production.
Run rapid prototyping and MVP delivery for client AI applications.
Establish practices for code quality, security, and AI data privacy compliance.
Manage cloud infrastructure and watch API usage so token costs stay predictable.

LangChainLlamaIndexOllamaPythonFastAPIPineconeAWSGoogle CloudDocker

2026 → Now

Operations Leader

Anchor Relationships, Portugal

The whole operations stack: lead to cash, signatures, payments, and scheduling, automated end to end.

Details

Architect and maintain the core operations stack, keeping data in sync across platforms.
Build lead-to-cash workflows that stream website data into CRM and marketing systems.
Integrate electronic signature and document management into onboarding and contracts.
Connect billing events directly to client accounts and communication channels.
Configure cloud infrastructure, domain routing, and web security for customer-facing services.

TypeScriptCloudflareZapierAWSAzureDocker

2025 - 2026

Full-Stack Developer & Team Leader

Cyberr, Luxembourg

Built a platform connecting cybersecurity professionals with businesses, from an empty repository to 100,000 registered users.

Details

Designed and developed the full-stack platform from scratch within six months.
Kept the system stable and scalable past 100,000 registered users.
Cut page load times by roughly 30% across the Next.js and TypeScript frontend.
Built and maintained secure backend services on GraphQL, PostgreSQL, and Node.js.
Deployed through Docker and Azure pipelines.

Next.jsTypeScriptGraphQLPostgreSQLNode.jsAzurePlaywright

2023 - 2024

Full-Stack Developer & Team Leader

ITGest, Portugal

Led a small team delivering enterprise applications across logistics, energy, and transport.

Details

Managed a small development team and shipped enterprise-grade applications.
Built interfaces in Vue, React, and Angular against real operational workflows.
Maintained and extended APIs on NestJS and Express.
Integrated systems across MySQL and MongoDB.
Introduced CI/CD, automated testing, and code review as standard practice.

ReactVueAngularNestJSExpressMySQLMongoDB

2021 - 2023

Product Developer & Researcher

Universidade de Aveiro, Portugal

Deep learning driver monitoring for AgroSafeBox: 96% accuracy on drowsiness and distraction, under 150 ms inference on target hardware.

Details

Developed the deep learning driver monitoring system behind AgroSafeBox.
Reached over 96% accuracy detecting drowsiness and distraction.
Brought model inference latency under 150 ms on target hardware.
Built the data pipelines and trained the models in PyTorch, OpenCV, and TensorFlow.
Shipped a real-time monitoring dashboard in React.

PythonPyTorchTensorFlowOpenCVKerasC / C++React

2020 - 2021

Junior Data Analyst & Developer

Stats Perform, Portugal

Real-time sports data collection and quality assurance, plus the dashboards that made the data readable.

Details

Contributed to real-time sports data collection and quality assurance.
Built dashboards consuming REST APIs.
Analysed match data and traced issues back through game-related tickets.

Power BITypeScriptReact

The stack.

AI & machine learning

LangChainLlamaIndexOllamaAnthropicMCPPineconePyTorchTensorFlowKerasOpenCVscikit-learnPandasFastAPI

Web3 & blockchain

SolidityEthereumHardhatWeb3.js

Frontend

ReactNext.jsVueNuxtAngularReact NativeSwiftTypeScript

Backend

Node.jsNestJSExpressDjangoFlaskGoJavaPHPGraphQL

Data

PostgreSQLMySQLMongoDBRabbitMQ

Infrastructure

AWSAzureGoogle CloudDockerLinuxNginxJenkinsCloudflare

Testing & authentication

JestPlaywrightKeycloakOAuth 2.0

Things I built.

CS Fundamentals

MDX

Computer science explained visually, with every complexity bound derived instead of asserted. Each topic also covers when the structure is the wrong choice and how it fails.

Binary search over 16 sorted keys. The bright cell is the current probe, the lit band is what is left to search.

cs-fundamentals

Auto Calendar

MCP server

An MCP server that reads work items out of a GitHub project and writes them into Google Calendar as scheduled blocks. Claude plans my week without me opening either app.

One working week. Hatched cells are already booked, solid cells are work items the server placed in what was left.

auto-calendar-python

Nextis Labs

Client work and products from my own company. RAG systems, agents, and internal tools, with more shipping through the year.

Company portfolio

miband-HR-pythonLive heart rate off a Mi Band 6 or 7 over Bluetooth.17 tiny-ECDH-pythonElliptic-curve Diffie-Hellman key agreement, small enough to read in one sitting.6 driving-monitor-pythonThe face tracker behind the research: drowsiness, fatigue, and distraction in real time.5 prod-blueprint-typescriptThe TypeScript starting point I clone whenever something has to reach production.TypeScript