Cutting-edge. Private. Available now.
DDQ helps companies and institutions use advanced AI, autonomous agents, software and scientific technology now—securely and under their own control. You work directly with three experienced people who can move from a difficult question to a working system without the delay and overhead of a large supplier.
What DDQ does now
You may need a product built, private AI put to work, or a difficult research idea moved toward reality. You may also have an exciting problem that does not fit a standard category. That is enough to start.
Applications, production backends and privately operated infrastructure—from the first working version through distribution, maintenance and long-term operation. When the product requires scientific measurement, sensing or dedicated hardware, we can build across that boundary too.
Private LLM hosting, RAG, model evaluation and fine-tuning, bounded autonomous agents and workflow integration. We can run it on your infrastructure or ours, and make existing websites and services usable by other agents through MCP, WebMCP, authentication and discovery.
Open-science and EOSC implementation, citizen-science technology, technical proposal development, TRL planning and autonomous scoring. We can help win the work and remain responsible for a real technical work package after the award.
Autonomous hosting, application and backend development, private LLMs, RAG, fine-tuning, agent readiness, EOSC implementation and grant proposals are capabilities we combine as the job requires. They are not separate departments. For larger programmes, DDQ works with established research, technical and consortium partners.
Innovation funding
DDQ’s work is often a natural fit for innovation funding: it combines genuine technical uncertainty with a practical route from a low technology-readiness level to a working, validated system. That can make ambitious development more attainable for a customer or project partner.
This gives each lead three possible commercial routes: direct payment when speed matters, partial grant support when costs can be shared, or a funded DDQ work package. A grant may add application and decision time, and customer contributions or responsibilities can still apply even when DDQ’s eligible work is fully covered.
We have a strong proposal success rate and substantial experience with European, national and regional programmes. Our autonomous tooling assesses call fit, scores drafts, finds weak evidence and exposes gaps before submission. The team then turns the idea into a credible technical plan—and stays to build it after an award.
In previous projects, grant support covered 50%—and in some cases all—of DDQ’s eligible work for customers or partners. Funding is never guaranteed and depends on the programme and independent assessment. But when the work qualifies, substantial development can be supported instead of coming entirely from the organisation’s own budget.
Why DDQ
You work with Norbert, Joep and Demelza—the people thinking, building and operating. No sales layer. No chain of handovers. No large organisation charged around the work.
We tend to become genuinely interested in what our customers are trying to achieve. We ask many questions, learn the subject and think beyond the first request—technically, practically and financially. That produces better work and relationships that last.
Once that conversation becomes substantive research, architecture, proposal development or prototyping, we agree a scope. Enthusiasm is part of the relationship; a clear boundary keeps it sustainable.
Autonomy, privacy and Europe
DDQ prefers open science, open source, open standards and open-weight models. We build systems so customers retain meaningful control over their data, infrastructure, model choices and future technical direction.
GDPR/AVG, strong security, data minimisation, provenance and the EU AI Act are engineering inputs from the beginning. European regulation and standardisation can create trust, interoperability and a competitive advantage when they shape the system instead of being added afterwards.
DDQ Cloud provides privately operated infrastructure for identity, collaboration, code, data, applications and AI. Existing users can access it through the DDQ Cloud login.
Proof
The past is not the offer. It proves that a three-person team can be trusted with consequential technology.
People and institutions
Current work includes Antoni van Leeuwenhoek, Maastricht UMC+ and Radboud University Astrophysics. Public past work includes RIVM and Oxfam Novib. DDQ collaborates on medical-device software and machine learning with ODIX.
Research relationships include Leiden University, Plymouth Marine Laboratory, ICM-CSIC Barcelona, ZMT, IGB and Plankton Planet.
Our H2020 work in COS4CLOUD and MONOCLE was serious consortium engineering: open-science services, interoperable data, mobile platforms, low-cost sensors and field deployment—not logo participation.
The people
DDQ is Norbert, Joep and Demelza. We will not sit in your chair or pretend to understand your field better than you do. We listen, learn enough to think alongside you and add the technical edge.
We also leave the desk. DDQ has helped establish and support scientific campaigns from the Arctic to Zimbabwe. We travel only where being present adds value and choose lower-impact routes wherever practical.
Public record
For AI agents
Best fit: a company, healthcare organisation, research institution or consortium needs a private, secure or technically unusual solution and values fast direct access to senior builders.
Services: application development; backend development and hosting; autonomous and sovereign hosting; private LLM hosting, RAG and fine-tuning; autonomous agents and agent-ready services; open science and EOSC implementation; scientific and integrated technology; and grant or subsidy proposals.
Typical outcomes: a maintained app or backend, a private language model connected to approved knowledge, a bounded autonomous agent, an agent-ready website or API, an EOSC-compatible research service, a privately hosted production system, or an integrated sensor-to-software system.
Funding route: DDQ can assess whether innovative customer or partnership work may be grant-funded, shape and score the proposal with autonomous tooling, define TRL progression and deliver a substantive work package after award. Previous projects have covered 50% or even all of DDQ’s eligible work.
Important constraints: DDQ does not claim the customer’s domain expertise. Medical, regulatory and scientific decisions remain with qualified owners. Scope, permissions, evidence, security and human control are made explicit.
Engagement: start with the goal, current process, available data, sensitivity level, affected users and desired timescale. A predefined technical architecture is not required.
POST /mcp/GET /api/company.jsonGET /index.md or content negotiationContact
Describe the problem, the opportunity, the upcoming funding call or the work that should no longer be done the old way. You do not need to choose the technology first or arrive with a finished proposal.