DvarkAB

Projects

The kinds of work we take on.

Six engagement shapes we know well, each grounded in work we have already shipped. If yours does not fit any of them, that is usually a good sign: the unusual ones are our favourites.

Greenfield / platform build

Building a product’s entire backend from an empty repository

When the product exists as an idea and nothing else, we design and build the whole platform: services, data layer, deployment, observability and on-call, then operate it in production. Case in point: the backend for Ray, a gaming-first Chromium browser, built and run by a single Dvark engineer from the first commit to hundreds of millions of requests a month.

  • Greenfield
  • Cloud backend
  • Production ownership
0 → 100M+monthly requests · 500K+ users · one engineer

Native & browser engineering

Deep work inside browsers, engines and compilers

The unforgiving layer: C++ inside Chromium-based products, game engines, network protocol design and custom compilers. We have shipped C++ in a production browser codebase and built a protocol compiler with lower overhead and faster serialisation than protobuf and gRPC, used as the backbone of a studio’s whole platform.

  • C++
  • Chromium
  • Compilers
  • Protocol design
C++ / RustChromium internals · custom compilers

Architecture at global scale

Platform architecture where the blast radius is the whole business

Designing and reviewing architectures that serve entire industries: a casino platform delivering 3,000+ games from 50+ studios to 900+ operators across Azure, AWS, Cloudflare and on-premises data centres. Technology selection, standards, reviews and the hard trade-offs, made with the people who run it.

  • Multi-cloud
  • Hybrid
  • Data streaming
  • Standards
900+operators served across four infrastructures

Migration & modernisation

Moving production without stopping the business

Datacenter to cloud, cloud to cloud, and cloud back to on-prem when sovereignty or cost demands it. We led modernisation journeys for enterprises across banking, healthcare, e-commerce, gaming and space at AWS, and have moved a company’s entire production environment between clouds without a rebuild.

  • Cloud migration
  • On-prem
  • Hybrid
  • Cost
10+industries, from banking to space

WebAssembly & porting

Native code in the browser, WebAssembly on the server

Two sides of the same coin: WebAssembly on the server for fast, isolated and dense compute, and porting existing applications, games and engines from C++, C# or Rust to run in the browser. We have built runtimes, toolchains and a patented self-orchestrating cloud platform on exactly this foundation.

  • WebAssembly
  • Porting
  • Server-side Wasm
  • Patents
02granted patents in distributed cloud computing

Rescue & incident response

Called in when production is on fire

Outages, scaling cliffs and runaway cloud bills. At AWS our principal was called out to customers mid-outage to find the root cause and design the fix. The engagement ends with the incident resolved and the architecture changed so it cannot happen the same way twice.

  • Incident response
  • Performance
  • Cost control
RCAroot cause · stabilise · prevent

Next step

Recognise your project?

Tell us where you are: blank repository, scaling wall, or migration deadline. We reply within two working days.