Loading
A Node.js consulting and engineering practice: backend, web and mobile, in TypeScript on AWS serverless. We design, build and repair the platforms that are not allowed to fail: checkout, event pipelines, partner APIs. And we ship trilingual products in Arabic, English and French, full RTL included.
Taking on new engagements
Get a diagnostic within 48 hA written one-page diagnostic: risks, priorities, first step. Not a sales call.
You are starting from a blank page: we design the architecture, build it and take it through to production.
The system already runs and has a history: we stabilize it, modernize it and extend it.
We have done both across IT services, e-commerce, banking and financial services, and food service: first-product startups as well as established platforms.
Senior engineers on your critical path, not a body shop: the people who scope your mission are the people who write the code.
Have something that has to be right?
Get my one-page diagnosticTwelve services, each with its own detail page. Most engagements combine two or three of them.
Backend services in Node.js, TypeScript and NestJS, from the first endpoint to production operation.
Read the detailREST and GraphQL APIs that partners build against: contract-first, versioned, documented.
Read the detailServerless architecture consulting on AWS: we design, build and operate event-driven platforms with Lambda, Step Functions, SQS and EventBridge.
Read the detailWe take over existing backends and modernize them without stopping delivery.
Read the detailMeasured optimization of backends under real traffic: latency, throughput and cold starts.
Read the detailThird-party integrations that hold up in production: payments, delivery platforms, cloud vendors.
Read the detailFast, accessible, multilingual web apps and sites, including full Arabic RTL support.
Read the detailProducts that work in Arabic and in French or English, with real right-to-left layout rather than a mirrored afterthought.
Read the detailCross-platform apps with React Native, Expo and PWA, built on the mobile backend they depend on.
Read the detailCompany sites and online stores that load fast, can be found, and read correctly in every language you sell in.
Read the detailMonthly upkeep of production systems: updates, security patches, monitoring, incidents and a written report.
Read the detailLLM features and workflow automation built as backend systems: queues, retries, cost control and fallbacks.
Read the detailYou talk to the person who writes the code. We scope the work, design it, build it, and hand over the runbooks. Nothing gets passed to a stranger after the call: whoever you meet is who delivers.
We would rather tell you a piece of work is not a fit than take it and be wrong about it.
Recently: Lambda cold starts cut from about 8 s to 1.5 s; a cloud marketplace fully reindexed with zero service interruption. Every launch ships with runbooks and 30 days of post-launch support.
See how we engageThe stack behind the projects above, in production, not on a wishlist.
Have something that has to be right?
Get my one-page diagnosticIt starts with a free 30-minute scoping call and a written one-page diagnostic. From there we run either a fixed-scope project or an embedded engagement billed on time, always in Node.js, TypeScript and AWS serverless.
Remote first, with working hours overlapping your team, from Toronto to Doha. On-site days are possible in France (Paris, Marseille, Lyon, Lille).
Two formats: a fixed-scope project, with the scope and boundaries in writing before any code, or a an embedded engagement billed on time. The 30-minute scoping call is free and the quote carries no commitment.
Yes. Standard NDAs are fine, and we treat everything you share as confidential either way. It is also why you will not find client logos or case studies on this site: our clients work under NDA, and we keep their names out of our marketing.
Field notes on Node.js, TypeScript and AWS serverless, in the order we publish them.
Lambda has no timer of its own, so a scheduler invokes it: an EventBridge Scheduler schedule in 2026, with timezone-aware cron, one-off at() runs and a retry policy per schedule, or a legacy EventBridge cron rule that only understands UTC. Here is the SAM and SDK code for both, and the timezone, idempotency and monitoring details that decide whether the job survives.
Provisioned concurrency keeps initialized Lambda environments ready before requests arrive, and bills for them whether they serve traffic or not. It earns its keep on synchronous, user-facing paths with a real latency budget, and never on async or queue-backed work.
Yes, DynamoDB is serverless: nothing to patch, nothing to size, nothing to fail over. But serverless describes the operational model, not the billing model. Only on-demand capacity bills per request, and the data model, the partition key and the cost stay entirely yours.
AWS Firewall Manager is not a firewall: it is the policy manager that deploys and enforces WAF, Network Firewall, security group and Shield Advanced configurations across an AWS Organization. Here is which service inspects traffic, which one only governs, and what a single account actually needs.
AWS Batch is a compute scheduler for containerized jobs; Step Functions is a workflow orchestrator that holds state and handles failure. They solve different problems, and the usual answer on a real pipeline is a state machine that submits a Batch job and waits for it.
AppSync is a managed GraphQL endpoint billed per operation, with subscriptions, caching and Cognito or IAM authorization included. Apollo Server is a library you host on Lambda or Fargate, trading operational work for full control of the schema, federation and data sources.
Tell us where things stand today. We reply within 24 hours, and within 48 you get a one-page diagnostic: what is at risk, what is worth doing, and what we would do first.