FOUNDATION
Cloud Foundation & IaC
Accounts, networks, and Terraform your team can read, extend, and trust.
- AWS
- GCP
- Azure
- Terraform
DevOps · SRE · Platform Engineering
Cloud infrastructure that holds under load.
Walking Mountain builds, secures, and runs the platforms behind SaaS, fintech, and AI products — with reliability, cost, and compliance owned end to end. Based in China, working with clients at home and abroad.
elevation · reliability
15+ yrs
running production infrastructure
40%
cloud spend cut, without losing performance
$150K+
annual infrastructure cost removed
SOC-2
compliance programs shipped
What we do
Full-stack DevOps, from the account root to the on-call rotation.
Engage us for a single climb or the whole ascent. Every deliverable ships as code you own — no black boxes, no lock-in.
PLATFORM
Kubernetes Platforms
EKS and GKE clusters built to scale under real traffic and stay up.
- EKS
- GKE
- Helm
- Autoscaling
DELIVERY
CI/CD & Release Engineering
Ship faster and roll back safely — from semi-annual to on-demand releases.
- GitHub Actions
- Pipelines
- Self-hosted runners
RELIABILITY
Observability & SRE
Know before your users do. SLOs, error budgets, and on-call that works.
- Prometheus
- Grafana
- Loki
- SLOs
COST
Cloud Cost Optimization
Cut the bill and keep the performance, with the numbers to prove it.
- FinOps
- Right-sizing
- Spend attribution
SECURITY
Security & Compliance
SOC-2, hardening, secret rotation, and CVE remediation done properly.
- SOC-2
- WAF
- Secret rotation
- CVEs
MIGRATION
Cloud Migration
Move off Heroku, split the monolith, or cross clouds — without breaking.
- Heroku → cloud
- Monolith → containers
- Cross-cloud
AI-NATIVE
★
AI-Native Platform Engineering
Run LLM and agent systems in production — with the ops, cost control, and observability they demand.
- LLM ops
- Agentic systems
- MCP servers
- LangSmith
Where we go further than most
Selected outcomes
Numbers we have actually shipped.
From real engagements — client names withheld. The point is the shape of the result, and we can walk you through how each was done.
40% ↓
Cloud spend cut while migrating a platform off Heroku onto AWS with Terraform and a proper CI/CD pipeline.
- Migration
- FinOps
- Terraform
$150K+ / yr
Annual infrastructure cost removed on a multi-tenant SaaS — right-sized EKS, tuned RDS/ElastiCache, self-hosted Airflow (90% saving on that workload).
- Cost engineering
- EKS
- Airflow
Semi-annual → bi-weekly
Release cadence transformed with trunk-based development and robust CI/CD for a blockchain gaming platform.
- CI/CD
- Release engineering
$450 / day → controlled
Runaway LLM spend on a production AI SaaS traced and fixed via prompt-cache optimization and per-agent cost attribution.
- AI-native
- LLM ops
- Observability
How we work
Four steps up the mountain.
A clear path from “what’s on fire” to “your team can run this without us.”
- 01
Assess
We audit your infrastructure, cost, reliability, and security — and tell you what is actually broken, with evidence.
- 02
Design
A plan you own, written in code. Terraform, architecture, and docs — no proprietary layers, no lock-in.
- 03
Build
We implement alongside your engineers, not around them. Reviews, pairing, and knowledge transfer as we go.
- 04
Operate
Monitoring, on-call, and a clean handover so your team can run the platform with confidence.
The team
A senior team that has carried production infrastructure at altitude.
Walking Mountain is a partnership of senior DevOps and SRE engineers — and a team we are growing deliberately. You work directly with people who have owned production end to end, not juniors handed your account. Between us we cover complementary ground:
Infrastructure & AI-native platforms
Founding partner · Senior DevOps / SRE
Fifteen-plus years owning cloud infrastructure across SaaS, fintech, and blockchain — most recently sole infrastructure owner for a production AI SaaS. Re-platformed async workloads to Temporal, rebuilt observability on Prometheus/Grafana/Loki, and cut cloud spend at multiple clients.
- AWS + GCP
- Terraform · Kubernetes
- Temporal · Observability
- AI-native platform
Platform & cost engineering
Founding partner · Senior DevOps
Built and scaled cloud infrastructure from near scratch for a multi-tenant SaaS platform — templated Terraform for rapid client onboarding, EKS across dev/staging/prod, and SOC-2 Type 1. Removed $150K+ in annual infrastructure cost through right-sizing and self-hosting.
- AWS · EKS
- Terraform · CI/CD
- SOC-2
- Cost engineering
Based in Shenzhen · working across China and abroad · English & 中文
Get in touch
Tell us what’s on fire — or what you’re building next.
Whether it’s a 3 a.m. reliability problem, a cloud bill that keeps climbing, or a new platform you need built right the first time — we’d like to hear about it.