Daniel is currently unavailable
Daniel Sierra
Verified Expert in Engineering
Full-stack Engineer and Developer
Madrid, Spain
Toptal member since April 22, 2026
Bio
Daniel is a full-stack AI engineer and founder with more than 11 years of experience, passionate about building technology that delivers real value. He does not stop at strategy, but builds end-to-end AI products and helps businesses integrate AI in practical, meaningful ways. Daniel founded his own AI company, where he developed the entire tech stack from scratch, and brings additional depth through consulting across multiple industry sectors and teaching.
Portfolio
IE University
Python, Machine Learning, University Teaching, Artificial Intelligence (AI)...
Fitizens
Python, Artificial Intelligence (AI), Flutter, C++, Firebase, SQL...
Telefonica Tech
Python, Machine Learning, AI Consulting, Machine Learning Operations (MLOps)...
Experience
- Python - 11 years
- Artificial Intelligence (AI) - 11 years
- Data Science - 11 years
- Cloud Computing - 10 years
- Machine Learning Operations (MLOps) - 7 years
- Mobile Development - 5 years
- Flutter - 4 years
- Large Language Models (LLMs) - 2 years
Preferred Environment
Python, Artificial Intelligence (AI), Claude Code, Flutter, Firebase, Google Cloud, Large Language Models (LLMs), Data Science, Cloud Computing, Retrieval-augmented Generation (RAG), LangChain, Multi-agent Systems, LangGraph
The most amazing...
...achievement has been co-founding Fitizens, an AI fitness coach that analyzes user performance in weightlifting exercises from a single video.
Work Experience
Adjunct Faculty
2022 - PRESENT
IE University
- Delivered hands-on instruction to 300+ students across multiple cohorts, adapting technical content to a cohort coming primarily from non-technical undergraduate backgrounds.
- Taught Python for Data Analysis I and II, two core mandatory courses in the MBDS Master, covering Python fundamentals, advanced data manipulation with pandas, visualization, statistical analysis, and applied machine learning.
- Earned a Best Professor award at IE University, sustaining student evaluation scores consistently above 4.5 out of 5 across multiple cohorts of the MBDS program.
Technologies: Python, Machine Learning, University Teaching, Artificial Intelligence (AI), Data Analysis, Data Analytics, Decision Trees, NumPy, Pandas, GitHub, Data Quality, Git, Clustering Algorithms, Predictive Modeling, Cloud Platforms, Model Evaluation, ETL Development, LightGBM, XGBoost, Hyperparameter Tuning, Dashboards, Property Valuation, Streamlit, Data Architecture, REST APIs
Co-founder and CTO
2022 - 2026
Fitizens
- Architected and shipped a complete AI-powered video coaching product from scratch, spanning two Flutter mobile apps (iOS and Android), a Python/Firebase back end, and wearable integration.
- Engineered a multi-stage LLM video-analysis pipeline on Gemini 3 Flash to analyze videos from athletes executing weightlifting exercises.
- Designed and deployed a set of 30 on-device deep learning models to automatically count repetitions in real time of more than 100 from data extracted from IMU wearables.
- Built an end-to-end MLOps platform covering ETL pipelines, Label Studio annotation, experiment tracking, and hyperparameter optimization across a catalog of 500+ exercises.
- Architected a human-in-the-loop annotation pipeline with model-assisted pre-labeling.
- Gathered more than 40,000 exercise repetition IMU data and more than 900 videos of real users doing exercise.
- Directed 2 full technical pivots from B2B to B2C and from IMU-sensor classification to LLM video analysis, re-architecting the core stack and data model without disrupting production users.
- Automated release engineering with GitHub Actions and Fastlane, delivering multi-environment deployments across development, staging, and production to App Store and Google Play.
Technologies: Python, Artificial Intelligence (AI), Flutter, C++, Firebase, SQL, Large Language Models (LLMs), Claude Code, Gemini API, Data Labeling, AI Architecture, Engineering, Machine Learning Operations (MLOps), Data Science, Prompt Engineering, AI Agents, LLM Evaluation, Scikit-learn, MLflow, Mobile Development, Cloud Computing, Dart, Full-stack Development, Vertex AI, API Integration, Google Cloud, Data Analysis, Data Analytics, Data Engineering, Databases, RESTFul APIs, CI/CD Pipelines, Docker, Google Cloud Platform (GCP), NumPy, Pandas, PyTorch, Data Pipelines, Feature Engineering, Front-end, GitHub, GitHub Actions, Startups, Data Quality, Back-end, APIs, FastAPI, Git, Linux, Model Context Protocol (MCP), AI Tools, Decision Modeling, Spanish, Cloud Platforms, Model Evaluation, ETL Development, Data Modeling, Data Warehousing, A/B Testing, Optuna, XGBoost, Hyperparameter Tuning, OpenAI GPT-4 API, Convolutional Neural Networks (CNNs), Data Cleaning, Webhooks, Data Cleansing, Automation, Mathematical Modeling, Streamlit, OpenAI API, Whisper, Document Processing, Data Architecture, REST APIs, Workflow Automation & System Integration, Event-driven Architecture, Architecture, Agentic AI, AI Programming, AI Design, AI Integration, OpenAI, Zapier, LLM Reasoning, Full-stack, Workflow Automation
Lead Data Scientist
2019 - 2021
Telefonica Tech
- Led end-to-end delivery of B2B AI and ML projects across 6+ enterprise clients spanning energy, banking, transport, private security, and gambling sectors, owning scoping, modeling, deployment, and stakeholder communication.
- Designed and deployed an anomaly-detection system for smart-meter consumption data at an enterprise energy client, flagging irregular usage patterns including fraud, tampering, and faulty devices across a national meter fleet.
- Built per-location sales estimation models for the point-of-sale network of one of Spain's largest sports betting and gambling operators, producing revenue forecasts used for commercial planning and expansion decisions.
- Delivered a predictive-maintenance system for internal hardware assets at a major Spanish bank, combining sensor telemetry and maintenance log data to predict failures and prioritize field interventions ahead of time.
Technologies: Python, Machine Learning, AI Consulting, Machine Learning Operations (MLOps), Data Science, Scikit-learn, MLflow, Cloud Computing, Full-stack Development, Amazon Web Services (AWS), Azure, API Integration, Artificial Intelligence (AI), Data Analysis, Data Analytics, Data Engineering, Databases, Forecasting, ETL, Sales Forecasting, RESTFul APIs, CI/CD Pipelines, Docker, Decision Trees, NumPy, Pandas, Data Pipelines, Feature Engineering, GitHub, Data Quality, Back-end, APIs, FastAPI, Git, Apache Spark, Clustering Algorithms, Predictive Modeling, Recommendation Systems, AI Tools, Spanish, Cloud Platforms, Model Evaluation, ETL Development, Data Modeling, Data Warehousing, PostgreSQL, A/B Testing, Imbalanced-learn, LightGBM, Optuna, XGBoost, Hyperparameter Tuning, Natural Language Processing (NLP), Optical Character Recognition (OCR), Data Cleaning, Dashboards, Data Cleansing, Automation, Mathematical Modeling, Statistical Analysis, Data Architecture, REST APIs, Workflow Automation & System Integration, Server-side PDF Generation, Event-driven Architecture, Architecture, Workflow Automation
Senior Data Scientist
2018 - 2019
Telefonica Tech
- Deployed predictive maintenance and anomaly detection pipelines for a leading Spanish private-security company, surfacing early equipment failures and abnormal operational events across distributed infrastructure.
- Engineered demand forecasting and dynamic pricing models for ticket sales of a major Spanish transport operator, optimizing revenue per seat across routes and time windows.
- Spearheaded internal training and client-facing coaching on AI, ML, and its business applications, running workshops and mentoring junior data scientists on translating ML capabilities into operational outcomes.
Technologies: Python, Machine Learning, AI Consulting, Machine Learning Operations (MLOps), Data Science, Scikit-learn, MLflow, Spark, Cloud Computing, Full-stack Development, Amazon Web Services (AWS), API Integration, Artificial Intelligence (AI), Data Analysis, Data Analytics, Data Engineering, Databases, Forecasting, ETL, Decision Trees, NumPy, Pandas, Data Pipelines, Feature Engineering, GitHub, Data Quality, Git, Apache Spark, Clustering Algorithms, Predictive Modeling, Recommendation Systems, Prediction Markets, AI Tools, Spanish, Cloud Platforms, Model Evaluation, ETL Development, Data Warehousing, PostgreSQL, Imbalanced-learn, LightGBM, Optuna, XGBoost, Hyperparameter Tuning, Natural Language Processing (NLP), Rule-based Programming, Google BigQuery, Data Cleaning, Dashboards, Data Cleansing, Statistical Analysis, Azure Databricks, Data Architecture, Server-side PDF Generation, Energy, Graph Databases, Knowledge Graphs
Data Scientist
2015 - 2018
Synergic Partners
- Delivered a customer-segmentation platform for an international telecom company, profiling the national subscriber base to drive targeted commercial campaigns and churn reduction initiatives.
- Built predictive models and production-grade data pipelines for enterprise clients across telecom, energy, finance, and transport verticals as part of Synergic's cross-industry consulting practice.
- Designed and built internal tooling for large-scale data analysis on the Hadoop and Spark stack, enabling the team to process enterprise datasets and standardize the data-science workflow across client engagements.
Technologies: Python, Machine Learning, AI Consulting, Data Science, Scikit-learn, Spark, Cloud Computing, Amazon Web Services (AWS), Artificial Intelligence (AI), Data Analysis, Data Analytics, Data Engineering, Databases, Forecasting, ETL, Decision Trees, NumPy, Pandas, Data Pipelines, Feature Engineering, GitHub, Data Quality, Git, Clustering Algorithms, Predictive Modeling, Spanish, Cloud Platforms, Model Evaluation, Imbalanced-learn, LightGBM, Hyperparameter Tuning, Rule-based Programming, Data Cleaning, Data Cleansing, Statistical Analysis
Experience
Fitizens Athlete
https://fitizens.ioFitizens Athlete is a Flutter mobile app for iOS and Android that delivers real-time, professional-grade technique analysis for athletes and serious fitness enthusiasts. It combines Bluetooth-connected wearables with heart rate and 9-DOF IMU sensors, on-device video recording, and a multi-agent LLM video-analysis pipeline to automatically count repetitions, assess exercise form, and generate bilingual per-rep coaching feedback from a single phone recording.
The product ships with an extensive exercise catalog, a custom workout builder (EMOM, AMRAP, time-based, and rep-based sessions), weekly and monthly progress dashboards with HR zones and muscle activation views, and social sharing of technique clips.
I built the solution on Flutter, Riverpod state management, Firebase back end (Auth, Firestore, Storage, Crashlytics, Remote Config), Go Router navigation, and a privacy-first architecture in which raw video is preprocessed locally using C/C++ libraries before any cloud analysis.
It was shipped to App Store and Google Play with 200+ users, built end-to-end by a two-person founding team with no external funding.
The product ships with an extensive exercise catalog, a custom workout builder (EMOM, AMRAP, time-based, and rep-based sessions), weekly and monthly progress dashboards with HR zones and muscle activation views, and social sharing of technique clips.
I built the solution on Flutter, Riverpod state management, Firebase back end (Auth, Firestore, Storage, Crashlytics, Remote Config), Go Router navigation, and a privacy-first architecture in which raw video is preprocessed locally using C/C++ libraries before any cloud analysis.
It was shipped to App Store and Google Play with 200+ users, built end-to-end by a two-person founding team with no external funding.
brainkeeper
https://github.com/dasirra/brainkeeperbrainkeeper is an open-source standard and Model Context Protocol (MCP) server that lets AI agents safely read and write a structured Markdown "second brain" vault. It defines a spec for PARA-style folders, YAML frontmatter, tags, and lifecycle rules, then enforces that contract at the protocol boundary so notes stay consistent no matter how the agent's session went.
I designed the spec, built the Python package (core engine, FastMCP server, and CLI), and shipped it to PyPI as `brainkeeper`. The server exposes thirteen tools across primitive, conventional, and semantic layers, backed by a live in-memory index that keeps lookups fast as vaults grow past thousands of notes. Atomic writes prevent the agent from clobbering edits made in Obsidian or VS Code.
The stack includes Python 3.11+, FastMCP, pydantic, Ruff, and GitHub Actions CI. The package is built with `uv`, distributed via `uvx`, and released under the MIT License.
I designed the spec, built the Python package (core engine, FastMCP server, and CLI), and shipped it to PyPI as `brainkeeper`. The server exposes thirteen tools across primitive, conventional, and semantic layers, backed by a live in-memory index that keeps lookups fast as vaults grow past thousands of notes. Atomic writes prevent the agent from clobbering edits made in Obsidian or VS Code.
The stack includes Python 3.11+, FastMCP, pydantic, Ruff, and GitHub Actions CI. The package is built with `uv`, distributed via `uvx`, and released under the MIT License.
BareTimer: A Timer for Your Workouts
https://baretimer.comA Flutter-based app for CrossFit and interval training, shipping on iOS and Android from a single codebase. Built to start any timer in under 10 seconds, with full offline support and no ads, accounts, or tracking.
Engineering: Clean Architecture with Riverpod for state management and a pure-Dart timer engine that is fully unit-tested. Five modes (Countdown, For Time, AMRAP, EMOM, Intervals) share one declarative tick function, so behavior stays consistent and easy to extend.
Quality and delivery: Flutter analyzes clean, 150+ unit and widget tests, Maestro UI flows, released to TestFlight and Google Play through fastlane lanes.
Engineering: Clean Architecture with Riverpod for state management and a pure-Dart timer engine that is fully unit-tested. Five modes (Countdown, For Time, AMRAP, EMOM, Intervals) share one declarative tick function, so behavior stays consistent and easy to extend.
Quality and delivery: Flutter analyzes clean, 150+ unit and widget tests, Maestro UI flows, released to TestFlight and Google Play through fastlane lanes.
Numio: Agentic RAG-powered Tax Assistant for Spanish Freelancers
Spanish freelancers (autónomos) often struggle with opaque and ever-changing tax rules, losing valuable time. Numio provides quick, plain-language answers to their tax and regulatory questions, each one backed by citations to official sources (BOE, DGT, AEAT). I built the entire system end-to-end, focusing on the following elements:
• Developed a production-grade RAG pipeline that scrapes, chunks, and embeds official sources (BGE-M3) into PostgreSQL using pgvector, orchestrated by Dagster for efficient ingestion-to-index flow.
• Implemented a hybrid retrieval system that combines keyword (BM25) and semantic search. This is further refined by a neural reranker, multi-query expansion, and contextual compression to provide precise and relevant context.
• Utilized Gemma4 31B for grounded generation, ensuring every claim can be traced back to a cited source.
• Applied a rigorous quality control framework, featuring an automated evaluation harness with RAGAS, custom citation validation, ablation studies, and full observability through self-hosted Langfuse.
• Offered real-world user interfaces via a Streamlit web app and a live Telegram bot.
The system is built in Python with LangChain and fully containerized using Docker Compose.
• Developed a production-grade RAG pipeline that scrapes, chunks, and embeds official sources (BGE-M3) into PostgreSQL using pgvector, orchestrated by Dagster for efficient ingestion-to-index flow.
• Implemented a hybrid retrieval system that combines keyword (BM25) and semantic search. This is further refined by a neural reranker, multi-query expansion, and contextual compression to provide precise and relevant context.
• Utilized Gemma4 31B for grounded generation, ensuring every claim can be traced back to a cited source.
• Applied a rigorous quality control framework, featuring an automated evaluation harness with RAGAS, custom citation validation, ablation studies, and full observability through self-hosted Langfuse.
• Offered real-world user interfaces via a Streamlit web app and a live Telegram bot.
The system is built in Python with LangChain and fully containerized using Docker Compose.
Education
2013 - 2015
Master's Degree in Computer Science: Multimedia and Communications
University Charles III of Madrid (UC3M) - Madrid, Spain
2013 - 2015
Master's Degree in Telecommunication Engineering
University Charles III of Madrid (UC3M) - Madrid, Spain
2009 - 2013
Bachelor's Degree in Telecommunications: Telematic Engineering
University Charles III of Madrid (UC3M) - Madrid, Spain
Certifications
JULY 2021 - JULY 2024
PMI Agile Certified Practitioner (PMI-ACP)
Project Management Institute (PMI)
JUNE 2020 - PRESENT
Certified Scrum Master
Scrum.org
JUNE 2020 - PRESENT
Certified Scrum Product Owner
Scrum.org
MAY 2020 - PRESENT
ThePowerMBA
ThePower Business School
Skills
Libraries/APIs
Scikit-learn, NumPy, Pandas, REST APIs, Imbalanced-learn, XGBoost, OpenAI API, PyTorch
Tools
GitHub, Claude Code, Git, Whisper, Fastlane, Zapier
Languages
Python, SQL, Dart, C++
Frameworks
Streamlit, Flutter, Spark, LightGBM, Optuna, LangGraph, Apache Spark, FastMCP
Platforms
Firebase, Vertex AI, Amazon Web Services (AWS), Docker, Google Cloud Platform (GCP), Linux, Azure
Paradigms
Scrum, Agile, Mobile Development, ETL, Model Context Protocol (MCP), Rule-based Programming, Automation, Event-driven Architecture
Storage
Google Cloud, Databases, Data Pipelines, PostgreSQL, Graph Databases
Other
Artificial Intelligence (AI), Machine Learning, AI Consulting, API Integration, Data Science, Prompt Engineering, Data Analysis, Data Analytics, Decision Trees, GitHub Actions, Startups, Data Quality, Spanish, Hyperparameter Tuning, Data Cleaning, AI Integration, Large Language Models (LLMs), Engineering, Time Series, Digital Signal Processing, Gemini API, Data Labeling, AI Architecture, Business, Machine Learning Operations (MLOps), AI Agents, LLM Evaluation, MLflow, Cloud Computing, University Teaching, Full-stack Development, Data Engineering, Forecasting, Sales Forecasting, RESTFul APIs, CI/CD Pipelines, Feature Engineering, Back-end, APIs, Clustering Algorithms, Predictive Modeling, Recommendation Systems, Cloud Platforms, Model Evaluation, ETL Development, Data Modeling, Data Warehousing, A/B Testing, OpenAI GPT-4 API, Data Cleansing, Statistical Analysis, Document Processing, Data Architecture, RAG Architecture, Workflow Automation & System Integration, Server-side PDF Generation, Vector Databases, Architecture, Agentic AI, AI Programming, AI Design, OpenAI, LLM Reasoning, AI Agent Orchestration, AI Automation, Full-stack, Workflow Automation, Computer Vision, Front-end, FastAPI, Prediction Markets, AI Tools, Decision Modeling, Natural Language Processing (NLP), Optical Character Recognition (OCR), Convolutional Neural Networks (CNNs), Retrieval-augmented Generation (RAG), Google BigQuery, Dashboards, Webhooks, Property Valuation, Mathematical Modeling, LangChain, Azure Databricks, Multi-agent Systems, Dagster, Telegram Bots, Embedding Models, Scalable Vector Databases, Energy, Knowledge Graphs
Collaboration That Works
How to Work with Toptal
Toptal matches you directly with global industry experts from our network in hours—not weeks or months.
1
Share your needs
Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2
Choose your talent
Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3
Start your risk-free talent trial
Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.
Top talent is in high demand.
Start hiring