ushift

BACKGROUND

Please describe your company and position.

I am the CTO of Mirlo

Describe what your company does in a single sentence.

Mirlo is a  mobile virtual network operator that, via APIs, enables companies to seamlessly manage their SIM and eSIM lines, and connectivity services in an automated, integrated platform.

OPPORTUNITY / CHALLENGE

What specific goals or objectives did you hire ushift to accomplish?

• Automate and optimize our CI/CD pipelines to speed up development and deployment cycles.
• Architect and implement a scalable, secure AWS infrastructure (EKS, RDS, Redis, S3, EC2, Lambda) to support rapid growth.
• Integrate AI capabilities via Amazon Bedrock and build a full observability stack with Grafana and OpenTelemetry for real-time monitoring and improved reliability.

SOLUTION

How did you find ushift?

Referral

Why did you select ushift over others?

• Great culture fit
• Good value for cost
• Referred to me

How many teammates from ushift were assigned to this project?

2-5 Employees

Describe the scope of work in detail. Please include a summary of key deliverables.

We engaged ushift to design, build and operate our entire cloud infrastructure from the ground up. The main deliverables included:

CI/CD Automation: Implemented fully automated build-and-release pipelines (Git workflows, container builds, Blue/Green deployments) to accelerate feature rollouts and reduce manual errors.
Microservices Architecture: Collaborated with our dev team to design modular, fault-tolerant microservices, defining service boundaries, API contracts and IaC templates (Terraform & CloudFormation).
AWS Infrastructure Provisioning & Management:

EKS: Cluster setup, autoscaling, and secure networking.
RDS & Redis: Highly available databases with backup/recovery policies.
S3 & EC2: Storage, static assets hosting, and compute fleet management.
Lambda: Serverless functions for glue logic and event-driven jobs.
AI Integration: Integrated generative AI capabilities via Amazon Bedrock, including model selection, prompt orchestration and API wrappers for downstream services.
Observability Stack: Deployed a full monitoring and tracing solution using OpenTelemetry, Prometheus, Loki and Grafana dashboards—complete with alerting rules and run-book documentation.
24/7 SRE Support: Established an on-call rotation, incident response playbooks and automated health checks. Configured RBAC policies, encryption-at-rest/in-transit and a WAF to meet our security and compliance requirements.
This end-to-end engagement delivered a resilient, scalable platform that kept pace with our rapid growth and high-availability SLAs.

RESULTS & FEEDBACK

What were the measurable outcomes from the project that demonstrate progress or success?

Deployment Velocity: Reduced average deployment time from ~2 hours to under 10 minutes, enabling daily releases instead of weekly.
Availability: Achieved 99.95% service uptime  with zero customer-facing outages in the last quarter.
Incident Response: Cut mean time to recovery (MTTR).
Cost Efficiency: Optimized AWS spend by ~20%
Observability Coverage: Instrumented 100% of microservices with OpenTelemetry, reducing undetected errors  and enabling proactive alerting.
AI Adoption: Integrated Bedrock to reduce llm latency 

Describe their project management. Did they deliver items on time? How did they respond to your needs?

ushift organized the work in two-week sprints, tracked tasks in a shared board (we used Jira), and gave us concise weekly updates. All major deliverables—CI/CD pipelines, EKS cluster setup, AI integration—landed on schedule. When priorities shifted or urgent issues came up, they adjusted the sprint plan within hours and kept us in the loop via Slack and email. Their clear status reports and quick responses made it easy to stay aligned without over-engineering the process.

What was your primary form of communication with ushift?

Virtual Meeting

What did you find most impressive or unique about this company?

What I found most impressive about ushift was their technical talent, specifically:

Rapid codification of requirements: They translated our infrastructure needs into version-controlled infrastructure-as-code templates (CloudFormation) covering VPCs, EKS clusters, RDS/Redis and networking in just a few days.
CI/CD as code: Built fully automated pipelines so every change went live reliably, without manual intervention.
Concrete observability: Instrumented each service with OpenTelemetry, hooked metrics into Prometheus/Loki, and delivered Grafana dashboards we actually use for triage.
Hands-on problem solving: When we hit cross-VPC networking issues or CI race conditions, they dug in, patched the config or code, and added tests so the problem never recurred.

Are there any areas for improvement or something ushift could have done differently?

No specific feedback. They delivered on time and met all our needs.