You will get AWS SageMaker Cost Audit — Find & Fix Overspending
Project details
Most ML teams overspend on SageMaker by 40–60% —
idle endpoints, over-provisioned instance families,
and unmonitored training jobs silently drain budgets
every month. I find exactly where your money is going
and tell you how to get it back.
What you get:
• Live cost dashboard with endpoint utilization,
training job analysis, and notebook waste findings
• Exportable PDF findings report with severity
rankings and savings estimates
• Single AWS account review (additional accounts available)
• 30-minute findings walkthrough call
Delivered in 5 business days (Optional Add-ons may require more time)
idle endpoints, over-provisioned instance families,
and unmonitored training jobs silently drain budgets
every month. I find exactly where your money is going
and tell you how to get it back.
What you get:
• Live cost dashboard with endpoint utilization,
training job analysis, and notebook waste findings
• Exportable PDF findings report with severity
rankings and savings estimates
• Single AWS account review (additional accounts available)
• 30-minute findings walkthrough call
Delivered in 5 business days (Optional Add-ons may require more time)
AI Development Type
Model Tuning, Software MaintenanceAI Tools
Amazon SageMaker, MLflowAI Development Language
PythonWhat's included $2,500
These options are included with the project scope.
$2,500
- Delivery Time 5 days
- Number of Revisions 1
- Model Documentation
Optional add-ons
You can add these on the next page.
Fast 2 Days Delivery
+$500
Findings presentation
(+ 1 Day)
+$250
Second AWS account
(+ 2 Days)
+$500
Remediation support
(+ 2 Days)
+$750Frequently asked questions
About Jose
AI Engineer | RAG, LLM Apps, Agents, AWS & MLOps
Las Vegas, United States - 10:16 pm local time
I’ve worked on enterprise engineering teams delivering production systems through Agile/Scrum environments, owning work across architecture, development, infrastructure, CI/CD, cloud platforms, security, reliability, and cost optimization.
Today, I focus on building production-ready AI applications and platforms, including RAG systems, LLM-powered applications, AI APIs, and cloud-native AI infrastructure.
My approach is shaped by production engineering experience: AI systems should not only work in a demo — they need to be secure, observable, scalable, reliable, and economically sustainable.
What I can help you with:
• RAG and enterprise knowledge applications
• LLM-powered applications and AI APIs
• Agentic AI workflows and tool integrations
• Python backend development with FastAPI and Django
• AWS AI infrastructure using Bedrock, SageMaker, ECS, EKS, and Lambda
• PostgreSQL, vector search, and document retrieval pipelines
• MLOps, CI/CD, Docker, Terraform, monitoring, and deployment automation
• AI performance, latency, and cost optimization
• Cloud and AI security using IAM, VPC networking, encryption, and least-privilege design
Recent work includes:
• Designed and implemented enterprise RAG and knowledge-retrieval workflows
• Built and deployed scalable Python-based AI backends
• Implemented cloud infrastructure and deployment automation on AWS and Azure
• Reduced infrastructure and inference costs through autoscaling, right-sizing, and architecture improvements
• Improved reliability, observability, and security of production cloud systems
Tech stack:
Python, FastAPI, Django, React, TypeScript, PostgreSQL, RAG, vector search, LLM APIs, AWS Bedrock, SageMaker, Lambda, ECS, EKS, Azure, Terraform, Docker, GitHub Actions
If you’re building an AI product, RAG platform, document-processing workflow, or LLM-powered SaaS application and need someone who understands both the AI layer and the infrastructure required to run it in production, I can help.
I’m also happy to review an existing architecture or implementation and identify practical improvements around reliability, security, performance, and cost.
Steps for completing your project
After purchasing the project, send requirements so Jose can start the project.
Delivery time starts when Jose receives requirements from you.
Jose works on your project following the steps below.
Revisions may occur after the delivery date.
Kickoff (Day 1)
30-minute call to confirm scope, priorities, and access requirements. You grant read-only IAM access using our setup guide — takes under 10 minutes to configure.
Environment Inventory (Day 1–2)
We inventory all SageMaker resources in your account — endpoints, training jobs, notebook instances, pipelines, and supporting compute. CloudWatch utilization data is pulled for the preceding 30 days.