Telecom
A consumer cloud sold through three operators
Millions of users served across three carriers
Read the case studyDevOps and platform engineering
DevOps consulting that scales your product to millions of users and cuts the cloud bill
We take over the infrastructure work your team is stuck on: Kubernetes, cloud migrations, CI/CD, gRPC and REST services, real-time systems, data pipelines and AI operations. Senior engineers only, hands-on with your team, and a cloud bill that goes down.
What you get
- A smaller cloud bill
We cut one client's yearly cloud spend by 4 times, and every engagement starts by finding what you pay for and do not use.
- Launches that hold
Products switched on for millions of subscribers across six mobile operators, with launch days run from a runbook instead of an incident channel.
- A team that ships
Pipelines, golden paths, dashboards and on-call your engineers own after the handover, built with them on your repositories.
Scope and price agreed after a discovery call.
Millionsof subscribers across six mobile operators
$400k → $100kone client's yearly cloud bill, cut by 4 times
11services, from Kubernetes to LLMOps
- CKACertified Kubernetes Administrator
- CKSCertified Kubernetes Security Specialist
Stack we run every day
KubernetesgRPCRESTTerraformAnsibleArgo CDFluxHelmPrometheusGrafanaElasticsearchRedisMemcachedAirflowAWS GlueS3
01Case studies
Work we can show
Each study says what was built, what changed and what carried over to the engagements since.
SaaS
One platform, several country brands, one codebase
One platform serving several country brands from one Terraform codebase
Read the case studyIndustry
Products sold through operators
Telecom products fail in public. Capacity planning, real-time infrastructure and release engineering for operators and the vendors who sell to them.
Read the industry page02Services
What we scale, migrate and run
Eleven services in four areas, delivered by the same senior team.
Scale
Clusters, gRPC and REST services, real-time fleets and the reliability practice that keeps them up when the users arrive.
Migrate
Monoliths into services, on-prem systems and large data sets into AWS, GCP or Azure, with a rollback that stays open.
Run
The platform and the pipelines that let a small team ship every day.
AI operations
Models, LLM applications and agents run as production services, and AI applied to operations itself.
03Process
How an engagement runs, and who runs it
- Discovery call
To understand the system and the deadline.
- Written assessment
Findings and a plan you can accept in full or in part.
- Implementation
With your engineers, on your repositories, in small reviewed changes.
- Handover
Runbooks, dashboards and alerts your team owns.
04Guides
Engineering guides
Long-form answers to the questions we get asked most, with the limits you hit and how to test the result.
3 min read
How do you scale WebSockets to millions of concurrent connections?
How to scale WebSocket servers to millions of concurrent connections: connection limits, fan-out, sticky routing, load balancers, Kubernetes and load testing.
Read the guide3 min read
How do you scale a WebRTC SFU on Kubernetes?
How to run WebRTC selective forwarding units on Kubernetes: host networking, UDP port ranges, TURN placement, pod placement, autoscaling signals and load testing.
Read the guide3 min read
How do you migrate petabytes from on-prem to the cloud?
A plan for moving very large data sets to AWS, GCP or Azure: inventory, transfer methods, integrity checks, cutover windows and the rollback you hope not to use.
Read the guide3 min read
Horizontal or vertical scaling: which does your microservice need?
When to add replicas and when to add resources: the limits, costs and failure modes of each, and how to tell which one your service needs.
Read the guideTell us what you are scaling
Send a short description of the system and the deadline. A senior engineer replies within two working days, and pricing follows a discovery call.
Talk to an engineer or email admin@kitelang.cloud