focus
Software Engineering
Design, build and evolve applications and distributed systems that hold up in production.
Our consultants work within development teams on architecture, code, testing, performance and release. Their role depends on the context, but the objective stays the same: evolve the system without losing sight of how it will be operated.
What we address
- Modernise a legacy system without interrupting service
- Break up a monolith that has become impossible to evolve
- Design a coherent distributed architecture rather than an accumulation of parts
- Take over an application base whose debt is blocking business change
- Meet performance, scalability or availability requirements
- Open a system through APIs without weakening its core
Types of problem
- Rebuild of a customer-facing application for a digital savings player
- Development on a risk calculation chain in a banking environment
- Progressive decoupling of a back office into services, with zero-downtime cutover
- Moving an overnight batch process to an event-driven chain
Ecosystems
Java, Kotlin, Spring Boot · C#, .NET · Python · TypeScript, Node.js · React, Angular, Vue.js, Nuxt · microservices, DDD, hexagonal architecture · REST, gRPC, GraphQL · Kafka, messaging, event-driven
focus
Data & Platforms
Make data usable at scale, from the pipeline through to the use case.
A data platform is only worth what you can do with it. We build ingestion and transformation chains that are reliable, tested and observable, sized for real volumes rather than for demo day.
What we address
- Industrialise hand-built processes that no longer scale
- Absorb growth in volume or frequency
- Unify heterogeneous sources without building a machine nobody can maintain
- Make pipelines reliable when nobody knows why they fail
- Move from batch to streaming when the business requires it
- Produce regulatory reporting that is reliable and traceable
- Regain control of platform costs that are drifting
Types of problem
- Building new indicators on a high-volume counterparty risk calculation chain, with Spark
- Progressive migration of market risk calculations from ActivePivot to Spark, to reduce dependency on a proprietary OLAP engine without degrading performance
- Migration of an ageing Hadoop platform to a lakehouse architecture
- Industrialisation of regulatory reporting that was still semi-manual
Ecosystems
Apache Spark · Kafka · Databricks · Snowflake · BigQuery · dbt · Airflow · lakehouse, data lake, streaming · PostgreSQL, MongoDB, Cassandra · Power BI, Tableau, Looker
focus
AI Engineering & GenAI
Bringing AI into systems that are useful, measurable and under control.
AI is a priority investment area for Modders. We use it daily in our own development, research, documentation and automation workflows, and we support our consultants in mastering it. For clients, our positioning targets the move from experimental use to systems that are integrated, evaluated and operable.
The subjects we are investing in
- Bringing AI into an existing application or process without rewriting the whole system
- Building RAG chains and assistants connected to business data, with access rights handled properly
- Evaluating answers and behaviour: test sets, version comparison, regression testing
- Orchestrating workflows or agents with human validation at sensitive steps
- Observing and controlling cost, latency, quality, drift and incidents
- Automating repetitive tasks while keeping human responsibility for the result
AI in our own daily work
At Modders, AI is not only a commercial subject. Our people are encouraged to use it to speed up what can be sped up: development, documentation, research, automation. With one simple rule: understand, verify and stay accountable for what is produced.
Ecosystems
Python · LangChain, LangGraph, LlamaIndex · MCP · vector databases (pgvector, Qdrant, Milvus) · AWS Bedrock, Azure AI Foundry, Vertex AI · OpenAI, Anthropic and Mistral models, open-weight models · PyTorch, Hugging Face · MLflow · evaluation and observability for AI systems
focus
Cloud & Platforms
Give applications a runtime environment that is reliable, observable and under control.
Cloud solves nothing on its own: used badly, it reproduces the limits of the existing setup and bills more for it. We build technical foundations that development teams can use without becoming infrastructure experts.
What we address
- Migrate to cloud without reproducing the limits of the existing setup
- Shorten the time between a commit and a production release
- Make a platform observable: know what is happening before the incident
- Meet availability and recovery requirements
- Standardise environments without slowing teams down
- Regain control of infrastructure costs that are drifting
Types of problem
- Shared Kubernetes foundation and the delivery pipeline that goes with it
- Bringing a manually managed infrastructure under infrastructure as code
- Observability that is actually used, rather than dashboards nobody looks at
We bring these skills in when reliability, observability, industrialisation or running costs become structural constraints on the product.
Ecosystems
AWS, Azure, GCP · Kubernetes, Docker · Terraform, Ansible · GitLab CI, GitHub Actions, Jenkins · Prometheus, Grafana, OpenTelemetry
