Available for remote contracts · Targeting Canada
EDSONFLORES
MLOps Engineer·AWS · CDK TypeScript · SageMaker
5+Years AWS
4AWS Certs
Scroll
02 / Technical Stack
The Arsenal
☁
AWS Services
SageMaker95%
Lambda / API Gateway95%
EventBridge / S390%
CDK - TypeScript85%
Bedrock / GenAI70%
EC2 / ECS Fargate50%
⚡
Engineering
MLOps / Pipelines95%
Python90%
IaC / CDK90%
Data Engineering75%
ML/AI Architecture Solutions75%
RAG / LLM Apps70%
Full stack
SageMakerBedrockCDKLambdaApiGatewayRekognitionComprehendLexTranscribeTextractEventBridgeS3SNSAthenaDynamoDBRDS PostgreSQLCloudFrontSSMCodePipelineCodeBuildCloudWatchEC2ECS FargateVPCSQSTransfer FamilyCloudFormationPythonLangGraphFastAPITypeScriptDockerGitLab CISFTPFTPGoogle Apps Script
03 / Experience
Career Timeline
Machine Learning Engineer
Price Lab
May 2024 — Present
Lima, Peru · Remote
- Engineered CI/CD pipeline for our machine learning workflows, integrating GitLab with CodePipeline + CodeBuild, reducing deployment cycle from 2 days → 30 minutes
- Architected decoupled data processing, uses EventBridge + SQS to trigger Lambda/SageMaker Processing Jobs for data transformations, improving system reliability and scalability, monitored with CloudWatch and custom dashboards
- Migrated more than 10 heavyweight Django/EC2 processes to serverless SageMaker Processing Jobs, improving backend responsiveness by 50%
- Developed 8+ automated ingestion pipelines: Email/SFTP/FTP/Web → S3 csv files → SageMaker Processing Jobs → RDS updates
AI Engineer
FreelancePart-time
Jun 2024 — Present
Part-time · Remote
- Deployed 2 production RAG chatbot MVPs using Claude 3.5 Sonnet via Amazon Bedrock, improving retrieval accuracy by 35%
- Architected serverless API infrastructure using CDK TypeScript with API Gateway + Lambda for document indexing and DynamoDB writes
- Built document processing pipelines supporting PDF, DOCX, TXT with chunking, embedding generation, and vector storage
Lead Data Scientist
Morris & Opazo
Jul 2023 — May 2024
Lima, Peru · Remote
- Led cross-functional teams on 2–3 concurrent enterprise ML implementations end-to-end
- Architected 10+ technical proposals for enterprise ML/AI solutions with feasibility, cost estimates, and deployment strategies
- Built conversational AI chatbot using Amazon Lex + Lambda + DynamoDB, later enhanced with Claude, cutting response time by 50%
- Obtained AWS Solutions Architect Associate, ML Specialty, and Data Analytics Specialty
Data Scientist
Morris & Opazo
Apr 2021 — Jun 2023
Lima, Peru · Remote
- Developed 5+ ML proof-of-concept models (classification, regression, computer vision) for enterprise clients
- Architected serverless solutions leveraging Rekognition, Textract, Comprehend, Transcribe with Lambda + S3 + DynamoDB
- Created QuickSight BI dashboards connected to Athena for stakeholder reporting and KPI tracking
Intern Data Scientist
Morris & Opazo
Agu 2020 — Mar 2021
Lima, Peru · Remote
- Learned and applied foundational data exploration and ML concepts, algorithms, and tools in recommendation systems
- Trained in AWS services including S3, Lambda, SageMaker, and QuickSight for data processing and visualization
- Built a computer vision proof-of-concept using Rekognition for PPE detection in images
04 / Projects
Built in Production
Core Product10+ clients served
SKU Matching Pipeline
6-stage ML pipeline matching competitor SKUs to client product catalogs at scale.
- bge-m3 embeddings → Claude on Bedrock LLM refinement
- SageMaker Processing Jobs on ml.m5.4xlarge
- EFS model caching, EventBridge + DynamoDB coordination
SageMakerBedrockbge-m3EFSDynamoDB
Infrastructure50% faster backend
Multi-Tenant Orchestration
Decoupled, event-driven data workflows with client_id injection — zero shared state.
- Hybrid scheduling: SNS → Lambda (real-time) + EventBridge crons
- Parameterized SageMaker jobs enable parallel client execution
- 50% backend responsiveness improvement over Django/EC2
SageMakerEventBridgeSNSLambdaCDK
DevOps2 days → 30 min
MLOps CI/CD Pipeline
GitLab + CodePipeline + CodeBuild: automated model deployments on every main branch commit.
- Stages: lint → test → SageMaker Job → integration → deploy
- Deployment cycle: 2 days → 30 minutes
- Full CDK TypeScript — no ClickOps
CodePipelineCodeBuildGitLab CICDKLambda
View on GitHubFreelance · AI+35% retrieval accuracy
RAG Chatbot MVPs
2 production RAG systems using Claude 3.5 Sonnet via Amazon Bedrock.
- 35% retrieval accuracy improvement via vector embeddings
- Multi-format support: PDF, DOCX, TXT with chunking strategies
- Serverless: API Gateway + Lambda + DynamoDB (CDK TypeScript)
BedrockLambdaDynamoDBCDKClaude 3.5
View on GitHubOpen Source
IaC SageMaker Pipelines
CDK TypeScript multi-stack for automated SageMaker ML pipeline deployment.
- Multi-tenant client deployments via configuration changes
- Environment-agnostic — dev/prod parity by design
- Reusable patterns for parameterized client workflows
CDK TypeScriptSageMakerCloudFormationS3
View on GitHub05 / Certifications
AWS Certified
☁
AIF-C01AWS Certified AI Practitioner
Active · 2026 – 2029
☁
SAA-C03AWS Certified Solutions Architect
Renewal in progress
☁
MLS-C01AWS Certified Machine Learning
Renewal in progress
☁
DAS-C01AWS Certified Data Analytics
Renewal in progress
☁
CLF-C02AWS Certified Cloud Practitioner
Renewal in progress
Verify on Credly: edson-jair-flores-davila
06 / nAIthan Agent
Ask About Edson
Powered by Claude · Knows everything about Edson's experience, projects, and availability
✦
nAIthan Agent
Online
✦
Ask me anything about Edson's experience, projects, or availability.
Enter to send · Shift+Enter for newline
07 / Contact
Let's Work Together
Available for remote contracts as MLOps, ML Platform, or Senior ML Engineer roles.
Available Now
Remote · Open to Canada relocation
Built with Next.js · Deployed on Vercel · Backed by AWS CDK TypeScript
© 2026 Edson Flores · Lima, Peru