Development
We build technology for complex businesses.
If you are starting out, we build your platform from the first line. If you already operate, we integrate with what you have and build what is missing. Either way, from the idea to production and inside your organization.
In review2
- Student-t VaR in the pricing library
- Scoped tokens for the MCP server
In progress3
- Nightly ledger reconciliation job
- Backfill positions from the core
- Audit log for model runs
- Lead time4.2 d
- Change failure rate3%
- Coverage87%
Every release traced from the issue to the deploy.
def portfolio_risk(pnl, model):- var = hist_var(pnl, level=0.99)+ var = model.var(level=0.99, horizon=10)+ es = model.es(level=0.975, horizon=10) audit.log("risk.var", var=var, es=es)
Pipeline
- Unit tests
- Security scan · 0 critical
- Model validation
- Approvals 2 of 2
Released v2.3.0 to client-org
audit.log(release)on every release we ship, so each change has an owner and a reason.
Build
The applications, environments and delivery platforms your teams work on every day.
Software DevelopmentApplications, APIs, platforms and enterprise software built to measure.
- Web applications
- APIs
- Internal tools
- Integrations
- Legacy replacement
- Discovery and architecture before the first line of code.
- Delivery in your repositories, with review and traceability.
- Documentation and handover as part of the scope, not an extra.
Custom Libraries & SDKsCustom libraries and SDKs that package your business logic so your teams can reuse it across every system.
- Python and other language packages
- Financial and risk calculations
- Data validation
- SDKs for APIs
- Semantic versioning
- Published to your private package registry, with versioning and release notes.
- Automated tests and usage documentation with every release.
- One implementation of each calculation, instead of copies that drift apart.
Cloud & InfrastructureCloud architecture, infrastructure as code, migrations and platforms that scale.
- Infrastructure as code
- Migrations
- Cost and capacity
- Resilience
- Observability
- Environments described in code and reproducible from zero.
- Migration plans with rollback and a defined cutover window.
- Cost, capacity and availability measured, not estimated.
Git & Engineering PlatformsWe design and manage the infrastructure behind modern software teams.
- GitHub Enterprise
- Organization management
- Repository governance
- Access control
- CI/CD
- DevSecOps
- Branch policy, mandatory review and segregation of duties by default.
- Pipelines with tests, dependency and security scanning before release.
- Access reviewed periodically, with an audit trail a supervisor can read.
Data & AI
Turning scattered records into decisions, and decisions into automation that keeps its evidence.
Data & AnalyticsData platforms, engineering, analytics, BI and data governance.
- Pipelines
- Warehouse and lakehouse
- BI
- Data quality
- Lineage and governance
- Every indicator traceable back to the system it came from.
- Quality rules and lineage documented with the model.
- Ownership defined per domain, so numbers have an owner.
AI & Intelligent AutomationAI applications, agents, automation and retrieval over your own information.
- Agents
- RAG
- Process automation
- Evaluation
- Guardrails and audit
- Runs inside your environment, with scoped credentials.
- Every output logged: prompt, sources, tool and result.
- Evaluated against a test set before it touches a real process.
Specialized technology
Where the domain is the hard part: quantitative finance, risk, and regulated legal processes.
Finance & Risk TechnologyQuantitative models, financial analytics, risk, investments and specialized platforms.
- VaR and Expected Shortfall
- Stress and scenarios
- Backtesting
- Portfolio analytics
- Model documentation
- Assumptions, limits and validation evidence ship with the model.
- Reproducible results: same inputs, same numbers, any day.
- Built to be explained to a risk committee and to a supervisor.
Legal & Regulatory TechnologyLegalTech, RegTech, compliance and automation of regulatory and legal processes.
- Contract workflows
- AML/CFT screening
- Regulatory reporting
- Obligation registers
- Evidence trails
- Rules written with the lawyers who will defend them.
- Evidence stored with each decision, not reconstructed later.
- Designed around the filing the regulator actually asks for.
The stack we build with
stack:
- cloud:
- [google-cloud,microsoft-azure ]
- gcp_services:
- [cloud-run,gke,bigquery,cloud-sql,pub-sub,cloud-functions,cloud-storage,vertex-ai,secret-manager,cloud-build ]
- azure_services:
- [aks,app-service,azure-functions,azure-sql,blob-storage,key-vault,entra-id,azure-openai ]
- languages:
- [python,sql,scala,go,kotlin,rust,r,typescript ]
- frameworks:
- [fastapi,django,pandas,polars,scikit-learn,pytorch,spring-boot,ktor,grpc,axum,tokio,tidyverse,shiny,react ]
- data:
- [postgresql,sql-server,bigquery,redis,apache-spark,dbt,apache-airflow,kafka,duckdb,parquet,pgvector ]
- platforms:
- [sap-s4hana,bloomberg,microsoft-365,power-bi,looker,databricks,snowflake ]
- delivery:
- [github,gitlab,bitbucket,github-actions,gitlab-ci,azure-devops,terraform,docker,kubernetes ]
- ai_platforms:
- [claude,openai,gemini,azure-openai,vertex-ai ]
- ai_open_models:
- [llama,mistral,qwen,gemma,deepseek,phi ]
- ai_tooling:
- [mcp,langgraph,llamaindex,hugging-face,vllm,ollama ]
Using something else? We integrate with what you already have.
Subscription service
Managed Git organizations
We take care of the governance, security and operation of your Git organizations, on their enterprise editions. The code stays yours; we keep things in order.
- GitHub Enterprise
- GitLab Premium · Ultimate
- Bitbucket Premium · Data Center
- Organization architecture, teams, roles and permissions
- Branch policies, mandatory review and CI/CD workflows
- Security monitoring and configuration audits
- Support with response times by severity
- Deployment control, as an optional service
- Periodic reports on the state of the organization
Ways of working
We choose the model together, based on how defined the scope is. All of them follow the same method.
Dedicated capacity
A team with the roles and dedication you need, for a monthly fee. We organize the work around your priorities.
When the product evolves continuously.
Hours bank
A set number of monthly or total hours, with time records each period.
When needs are variable or occasional.
Fixed price per deliverable
Scope, deliverables, schedule, payment milestones and acceptance criteria defined from the start.
When the scope is clear.
We also run proofs of concept and pilots, to validate an idea before committing to a full project.
What stays with you
Your code, in your organization
The code lives in your repositories from day one or, if it starts in ours, it is transferred with its history, issues and CI/CD workflows.
The rights to what we build for you
Work built specifically for you is assigned to your organization upon payment, and the tools of ours left installed keep working even after the service ends.
The knowledge
Technical documentation and handover sessions so your team can run and evolve what was built without depending on us.
Your data, out of any training
We never use the content of your data to train artificial intelligence models.
Our standardized services are contracted under a master agreement, the published terms of each service and a service order. Custom development projects are governed by a contract agreed with each client. See Legal
Book a consultation