Knowledge graph engineering
Full implementation on Neo4j, Amazon Neptune or Memgraph. Ontology design, RDF and semantic web, ETL pipelines, graph querying, visualisation and graph-native AI.
Independent consultancy · Brno, Czech Republic
Knowledge graphs, data solutions and custom web. Built properly.
Nodeworks is a small technical consultancy specialising in knowledge graphs, AI and data engineering. You work directly with the people building your project, from initial design through to delivery.
- Graph databases
- Data and AI pipelines
- GraphRAG and ontology
- Handcrafted web
- Central European time
Graph
specialist
specialist
Neo4j, Amazon Neptune and Memgraph: data modelling, ETL, querying, visualisation, and applications on the graph.
Full
implementation
implementation
From data model to ingestion pipeline to API to dashboard. The whole stack, not just advisory.
Independent
No agency markup. You work directly with the people doing the work, from brief to delivery.
Based in
Brno
Brno
Central European time, Czech business environment. Available for in-person meetings in the Czech Republic.
What we build
Three areas of work
Knowledge graph engineering is the core specialism. Data and AI, and web development, extend it into a complete technical offering.
Data, AI and search
Ingestion pipelines, OpenSearch and Elasticsearch with vector search, GraphRAG, MCP servers, AI agents, RAG pipelines, Power BI and Grafana dashboards.
Web and applications
Hand-written static sites, Next.js, Streamlit and e-commerce, with SEO, hosting and domain handling. Native Android apps and FastAPI backends.
How it works
Brief, proposal, build
Three stages. You know what is happening at each one, and nothing starts without a scope and a fixed fee agreed in writing.
01
Tell us what you need
A short email or a 30 minute call. No questionnaires. Tell us the problem, what you have tried, and roughly when you need it.
02
Receive a written proposal
Scope, deliverables, timeline and a fixed fee. If the project needs phasing, each phase is agreed and priced separately.
03
Build and deliver
Work proceeds against agreed milestones. You can see progress throughout, not just at the end. Delivery includes handover documentation.
About
The team
A small team of knowledge graph engineers and data specialists, based in Brno. Core expertise spans graph database implementation, ontology engineering and the semantic web, AI and data pipelines, GraphRAG, and LLM-integrated systems.
The independent knowledge graph consultancy space in the Czech Republic opened up in August 2026 when the leading specialist firm was absorbed by Neo4j. Nodeworks fills that gap as a vendor-neutral, independent practice.
Speaking
NODES 2026
Martyns is presenting at NODES 2026, the global knowledge graph conference hosted by Neo4j. NODES brings together graph practitioners, data engineers and researchers from across the industry.
Whether you want to connect around the conference, discuss a project or exchange ideas on knowledge graphs and AI, reach out directly.
Writing
From the blog
Technical writing on knowledge graphs, AI, semantic web and data engineering.
· Knowledge graphs
NaijaHealth: a graph-native health intelligence agent for Kano State
A knowledge graph over Nigeria's NHMIS data for all 44 Kano LGAs with an AI agent on top. Dual-grain graph model, idempotent DHIS2 ingestion, and nine agent tools across Cypher templates, text2cypher and vector search.
· Speaking
NODES 2026: Graph-Aware Recommendations with Neo4j GDS and Random Forest
In November Martyns is presenting at NODES 2026. The talk covers building a production-grade technographics recommendation engine using FastRP embeddings, graph-aware hard negative sampling and Random Forest. OOB AUPRC 97.3%.
Common questions
Frequently asked
Primarily knowledge graph and data engineering work for businesses that need production-grade implementations rather than a proof of concept. We also take on web and application projects, typically for clients who already trust us on the data side.
We do not take on projects where the main deliverable is a report or slide deck. Implementation and working software only.
A knowledge graph is a database that stores data as entities and the relationships between them, rather than rows in a table. It is the natural data model for problems involving connected data: recommendation engines, entity resolution, supply chain tracing, compliance graphs, network analysis, and any question that involves following chains of relationships.
If your current database struggles with multi-hop queries, or you are spending more effort maintaining join tables than on your actual product, a knowledge graph is worth looking at. A scoping call will tell you quickly whether the technology fits your problem.
Yes. Most work is remote by default. We are available in Central European time, which overlaps well with Western Europe and is workable for shorter windows with East Coast US clients.
For Czech clients who want in-person sessions, Brno and Prague both work. The Czech Republic, Slovakia and Austria are easy day trips.
By fixed fee for defined deliverables, agreed before any work begins. The fee is based on scope and complexity, not a quoted hourly rate applied post-hoc.
For ongoing work, a monthly retainer is available. After-support (maintenance and incident response) is on a separate, clearly priced arrangement.
Nothing is priced on this website. The price is in your proposal, after a brief scoping call.
Ready to start?
Tell me what you are trying to build. I will reply within one business day.
Get in touch