Product engineering
New applications and features, from early architecture and interfaces through APIs, data models, and production delivery.
- Web and mobile applications
- Backend services and APIs
- Product architecture and technical direction
Product & systems engineering
We design and build software products, platforms, and the systems behind them, from interface to infrastructure.
Product engineering is what we do. Systems thinking is how we do it.
What we do
What we build
We can take responsibility for a complete product or join around a difficult part of one. The work spans four connected areas.
Platforms & distributed systems
The shared capabilities a product needs as it grows, from multi-tenant foundations to event-driven services and integrations.
- Application platforms and internal tooling
- Event-driven and distributed services
- Integrations, permissions, and multi-tenancy
AI workflows & agents
AI-enabled systems that use models for judgment while keeping execution, state, and control explicit.
- Agentic workflows and tool integrations
- Evaluation, tracing, and model boundaries
- Human approval where actions need oversight
Infrastructure & delivery
The foundations required to ship, operate, and evolve a product without making infrastructure the product.
- Cloud architecture and infrastructure as code
- CI/CD and repeatable environments
- Observability, security, and developer tooling
How we build
Systems thinking, throughout
We treat product design, application architecture, AI workflows, and infrastructure as one connected system. Ownership, state, failure, and recovery are considered early, not left for production to expose.
01
Clear boundaries
Explicit contracts, state ownership, and permissions keep the system understandable as the product and team grow.
02
Durable execution
Queues, idempotency, bounded retries, deduplication, and replay make important workflows safe to depend on.
03
Bounded AI
Models handle judgment where they help. Code, workflow constraints, evaluation, and human approval control what happens next.
04
Built to operate
Telemetry, recovery paths, and operational controls are part of the design, so the people running the system can see and change it.
Technologies
Tools we use
The tools, patterns, and platforms we use most often, with short notes on why we choose them.
Application stack
The layer clients see first. Every entry here is running in something we shipped or operate ourselves, and each one has a boundary where it stops paying.
Next.js, React Native / Expo, NestJS + 5 more
AI systems
AI features are software with a probabilistic component. These tools keep the probabilistic part bounded, traced, and swappable.
Bounded agents, Langfuse, Model gateways
Platforms
Where the workloads run. Most of what we deploy lands on Google Cloud and Cloudflare, with Vercel hosting the frontends that benefit from it.
Cloud Run, Kubernetes, Cloudflare Workers
Infrastructure as Code
Infrastructure as code keeps resources and changes reviewable, and makes environments easier to reproduce.
Terraform / OpenTofu, Ansible, Infracost
Delivery
A reliable delivery pipeline reduces the routine work and uncertainty around testing, deployment, and rollback.
GitHub Actions, GitLab CI/CD, ArgoCD + 3 more
Observability & incident response
Useful observability narrows the search during development and incidents by connecting service behavior to traces, metrics, and logs. This section also covers what happens after the alert fires.
OpenTelemetry, Google Cloud Observability, Grafana + Prometheus + 3 more
Security & access
Security controls are easier to maintain when they are part of the normal delivery workflow and enabled by default.
Workload identity, IAM as code, Secrets management + 4 more
Patterns
Ways of building rather than products you can buy. The entries in the other sections are how we implement them.
Event-driven architecture, Durable execution, Multi-tenancy + 3 more
How we engage
Meet the product where it is
01
Start something new
Take a product from an early idea through architecture, implementation, launch, and the first rounds of learning.
02
Evolve a platform
Add capabilities, clarify boundaries, or strengthen delivery as an existing product and engineering team grow.
03
Own a hard subsystem
Join around a difficult workflow, AI system, integration, backend service, or platform problem that needs focused ownership.
Tools & open source
Small systems, built in the open
Alongside our product work, we publish focused tools and reference implementations from the problems we explore.
file-guard
↗Per-process access control for credential files, implemented as a Rust FUSE daemon.
RustSecurityFUSE
aipm
↗A suggest-only AI work bot with durable execution, tool integrations, and human approval.
AICloudflareTypeScript
dns-auditer
↗Scheduled DNS and web security checks that manage their own GitHub Issues.
PythonSecurityGitHub Actions
ansible-github-runner
↗An Ansible setup for isolated, self-hosted GitHub Actions runner fleets.
AnsibleCI/CDLinux