TroidX
[ Technologies ]

What we build with.

We're stack-agnostic by principle and stack-opinionated in practice. The list below is what we actually reach for — chosen for problems we've seen, not for résumé padding. We won't take on a stack where nobody on the team has five years of production scars.

01

Languages

We pick languages for the problem. Backends usually run on Go or TypeScript; data and AI work runs on Python; mobile clients on TypeScript (React Native) or native Swift / Kotlin where the platform demands it.

GoTypeScriptPythonJavaSwiftKotlinRust
02

Frontend & Mobile

Next.js for production web, React for app shells, React Native for cross-platform mobile, native when performance ceilings or platform integration require it.

Next.js 14ReactReact NativeExpoTailwind CSSFramer MotionAstro
03

Backend & APIs

We default to typed RPC where the team owns both ends, REST where the consumer is external, GraphQL only when the federation actually earns its keep.

Node.jsHonotRPCExpressSpring BootFastAPIGraphQL
04

Databases & Storage

PostgreSQL until you can articulate why not. Time-series in TimescaleDB. Blob storage in S3-compatible buckets. Vector indexes in pgvector for almost everyone, dedicated vector DBs only at scale.

PostgreSQLTimescaleDBRedisMySQLMongoDBS3 / R2pgvectorElasticsearch
05

Streaming & Async

Kafka where the volume earns it. NATS or Redis Streams where it doesn't. Temporal for durable workflows. Cron is still a fine answer for many problems.

Apache KafkaNATSRabbitMQRedis StreamsTemporal
06

Cloud & Infrastructure

AWS, GCP, or Vercel — chosen for what the product actually needs, not what the platform vendor sponsored last conference. Infrastructure-as-code by default.

AWSGoogle CloudVercelCloudflareKubernetesDockerTerraformPulumi
07

Observability & Operations

OpenTelemetry as the substrate, with whichever backend the team prefers. Sentry for errors, PagerDuty for human paging, Grafana for dashboards.

OpenTelemetryGrafanaDatadogSentryPagerDutyHoneycomb
08

AI & ML

Frontier models where they earn their cost; smaller models or classical ML where they don't. Eval-driven development, regression tracks, cost dashboards in CI.

Anthropic ClaudeOpenAIMCPBraintrustLangSmithpgvectorHugging Face
09

Payments & Identity

Stripe where the geography supports it. Regional gateways and partner banks where it doesn't. Identity via Clerk or Auth0 for most teams; build vs. buy is decided per engagement.

StripeRazorpayClerkAuth0OktaFirebase Auth
10

DevOps & CI

GitHub Actions for most, GitLab when the rest of the org uses it. Preview environments per branch, blue/green or canary for production cutovers.

GitHub ActionsGitLab CIArgo CDRenovateTrunk

Working in something not on the list?

We'll tell you honestly whether we're the right team. If we're not, we'll point you at someone who is.

Tell us your stack
Strategy call