Logo pictureAbout Us

We are an AI-driven organization based in The Netherlands Amsterdam area, which promotes the use of linked open data platform(s) and Deep/Machine Learning solutions for smart home/office automation using Internet-of-Thing appliances that support programmatic/API controls. We provide top-of-the-line products & services packaged & customized to augment & integrate data streams in home/office context to make Internet-of-Things appliances&virtual assistants smarter through applied technologies from Artificial Intelligence, Machine Learning, Linked Open Big Data Science & Semantic Web. We provide Data Science & AI Solutions for recommenders, chatbots & digital assistants to acquire, integrate&combine knowledge from Internet-of-Things&Semantic Web repositories using Data Science pipelines, Machine Learning algorithms & models, to focus on smart applications like Object Recognition&Classification, Scene Understanding, Emotion Recognition&Content Summarization in Text&Multimedia content. Live with us our vision for Digital AI-enhanced Home/Office Ecosystem.

How it works? Instructables/NewsletterYou sign up with your email or SMS, city and usage scenario of interest, and we respond by e-mail with products & services are available in your area. If you opt-in for our newsletter, you get weekly top 10 package improvements for home/office automation. Products You get a hand-picked cost, efficiency & energy-efficient package of smart appliances, including sensors, actuators and IoT hubs/edge devices for your specific smart home/office scenario of interest. Services Our main added value is an assistant mobile/web app & appliance to control & customize the "brains" of your home/office on a local server (isolated from internet) or, if you so prefer, in a fully secured cloud. You have one week to one month to customize the package and try out the software and to decide whether you would like to rent or buy it, further customize it, or you want to stop using it altogether. You can stop at any time your subscription and return any rented products within the prepaid subscription costs. We are dedicated to protect your privacy & guarantee the safety of your sensitive data stored within our platform. Our data appliance software & hardware is periodically tested against data leaks and/or unwanted access. For cloud-connected appliances, you can opt in to participate in a device hunt to collectively harden the security of the platform. You make reversible steps, no unwarranted risks for your data or unnecessary up-front investments & hidden costs. You can choose between prepaid subscriptions or pay only for what you use. For an affordable subscription fee (ranging from 10 up to 20 EUR / month), we help you make use of and keep our supported devices up-to-date in a safe, smart & comfortable way. On request, we provide fast & responsible delivery, collection or transfer of ownership of consumables, data, instructables, software/models and services for the supported appliances. Note:We are still growing our business. We provide the products & advisory services across the world, but some rental, logistics, repair & delivery functions are for now only available within EU/Europe, until we have arrangements with an organization in your area.

Contact Details

Knowledge Flows
A: Amstelveen, The Netherlands
T: (+31) 020 823 2003
E: contact@knowledge-flows.com
W: www.knowledge-flows.com
T: twitter.com@flowsknowledge
G: github.com/KnowledgeFlows

Profile pictureAbout Me

I'm a Freelance Data Scientist, AI MLOps/DevOps, Solution Architect & incidental blogger on full-stack AI solutions, i.e. content augmentation & multi-modal interaction & task optimization within hybrid teams (people, devices & bots) using video, speech, gesture, text & sensor signals and applying best practices, models, algorithms & methods from Deep Learning/Machine Learning/Big Data/Edge-Fog-Cloud applied to Internet-of-Things. In recent past I have contributed to several enterprise IoT solutions for video situation understanding, language understanding and signal analytics from multimedia event streams and their integration with Data Science pipelines and Machine Learning models deployed in production. I published academic papers on linguistic patterns recognition in texts for automated tagging, text-to-model guideline operationalization & model building & critiquing, and ad-hoc decentralized resource optimization for task coordination in distributed teams. I am passionate about extracting new insights from datasets and investigating how zero-shot learning, incremental learning and federated learning can take place with private and virtualized data sets, for instance data store in a knowledge graph structure like the one in a blockchain or in a social network. I have full stack development experience including building cross platform applications and services.

Contact Details

Radu Serban


Vrije Universiteit Amsterdam

PhD Information Management & Software Engineering November 2002

At Vrije Universiteit Amsterdam, I studied Math & Computer Science mechanisms of privacy by design in social media & e-commerce environments, creating the building blocks of a computationally efficient, secure & safe personal assistant integrated with a collaborative system of smart proxies which uses Machine Learning & Natural Language Understanding & Social Network Analysis for understanding e-transactions, meta-knowledge extraction about exposed digital items & potential security/privacy threats; using a web-of-trust and collaborative filtering recommendations to create an early risk detection adaptive distributed system to protect individual assets & interests online across social networks.

VU Amsterdam & Polytechnical University of Timisoara

MSc in Computer Science September 1997

As Computer Scientist, I am educated in Project Management, Expert Systems, System Theory, Computer Networks, Databases, Data Structures, Math, Programming & Electronics. My EU grant for completing MSc/Master thesis at Vrije Universiteit Amsterdam produced a distributed conceptual search engine prototype using declarative programming (Prolog) & distributed computing (mobile agents, websockets C/C++, Java RMI) & NLP/NLU (language understanding).


Knowledge-Flows/AI-Fluent, Data Science Services

Freelance Data Scientist August 2015 - present

Key Responsibilities:

  1. Define product designs, marketing strategies & implementations of AI & Machine Learning services-based products, and accompanying mobile or web apps to serve commercial & individual consumers.
  2. Understand & formulate end users needs & expectations for Internet-of-Things, edge-fog-cloud computing solutions through technology stack selections, maturity assessment roadmaps building & priority setting.
  3. Envision applied (Deep) Machine Learning solutions, building blocks & integrated services.
  4. Create technical content like blog posts, pitches, webinars, developer tutorials, etc to communicate product value propositions.


Data Scientist July 2020 - Mar 2021

Key Responsibilities:

  1. Identify & extract features of accomodations with highly stable price bids to help classify situations of price change & predict change magnitutde.
  2. Apply Data Science model training using full-stack Jupyter, Docker, Kubernetes, Spark, Hive, Oozie, Tableau.
  3. Develop solutions to sample & predict price changes in hotel bids from millions of parallel time series.

Accenture BV

Data Science Team Lead September 2015 - February 2020

Key Responsibilities:

  1. Define Proof-of-Concepts for Data Science solutions supporting Analytics-as-a-Service in multiple domains - Finance, Telecom, Energy & Utilities, Product Retail & Distribution.
  2. Lead development & engineering teams advising clients on product selection, analytics strategy & design, ensuring that client feedback is at the forefront of our roadmap.
  3. Plan & execute implementations of Big Data, Machine Learning & Internet of Things Cognitive Services Proof-of-Technology at top 500 Netherlands companies.
  4. Participate in marketing activities, exhibitions, congresses & hackatons to build awareness & promote our offerings to our business partners, existing & prospective clients.
  5. Define & lead Machine Learning-related internships with innovative use cases for applied Data Science & Machine Intelligence research in business, in particular Internet-of-Things, Big Data & Cloud Data Science.
  6. Act as Thought Leader in Internet-of-Things, Big Data & Go-to-Market Digital Technologies community, giving presentations & hands-on labs that enable the prospective data science community to get maximum value from Machine Learning, Big Data & Internet-of-Things offerings.
  7. Design, implement, roll out & maintain a job recommender system for staffing candidates to roles & suggesting roles & courses to employees, product recognized through enterprise award in November 2019.

Systemation AES

Solution Architect April 2014 - August 2015

Key Responsibilities:

  1. Support pre-sales presentations with experiments and service architectures for Self-Service BI, Master Data Management, Big Data Analytics and Service Virtualization
  2. Implement service virtualization solution for largest Telecom provider in The Netherlands
  3. Design & implement custom solutions for Self-service Business Intelligence for Social Media Analytics

Almende BV

Research Scientist August 2011 - November 2012

Key Responsibilities:

  1. Contributed research deliverables, designs & prototype implementations for 3 fundamental research projects financed by European Union.
  2. Designed & implemented a software prototype for decentralized & ad-hoc resource management app in emergency response to calamities.
  3. Designed & implemented predictive solutions for energy consumption reduction in smart buildings and for optimalization of workload control in adaptive data centra.


Freelance Data Architect & Data Engineer January 2011 - August 2015

Key Responsibilities:

  1. Designed & implemented solution designs for e-commerce websites & business-to-consumer mobile applications
  2. Created solution offerings for Natural Language Understanding, Semantic Tagging and Data Enrichment in legislative text corpus.
  3. Implemented a cross-platform mobile application for energy consumption reduction.


IT Architect July 2007 - August 2011

Key Responsibilities:

  1. Designed & implemented solutions for major IBM partners in The Netherlands
  2. Implemented J2EE (Java-based) enterprise solutions for realtime finance analytics processing integrated with open source (UML, Java, Linux).
  3. Implemented a scalable real-time smart meter reading analytics solution for largest Energy provider in The Netherlands.
  4. Co-designed co-implemented a Document Management Solution for Douane/Tax Office in The Netherlands.
  5. Acquired IT Architect curriculum attestations for TOGAF IT Architect certification.
  6. On behalf of IBM IT Architecture Design Authority reviewed & approved solution risk assessments for several large projects.


For more details on my skills, please visit my LinkedIn page.

  • Machine Learning: Classification, Regression, Clustering, NeuralNets, ConvNets, GANs, AutoML, Automated Feature Extraction
  • Data Science: Jupyter*, R dplyr/tidy*/SparkR, Python Pandas/PySpark
  • Deep Learning: Pytorch/Tensorflow/Keras/Caffe
  • ML models for Video/Speech/Text/Signal Analytics
  • AI/ML Solution Architectures
  • Big Data: Hadoop/Spark/Oozie/Hive/Flume/Kafka
  • Cloud DevOps: AWS/GCP/Azure, Docker, Kubernetes, Ansible, Linux/Windows/iOS/Android.

Get In Touch.

For details about our projects, comments and/or inquiries, please use the form below.

Your message was sent, thank you!