About Nestor
Your business needs automation, scalability, or AI integration? I build end-to-end cloud and AI platforms that work in production — not slides.
I specialize in AWS-native architectures, Kubernetes, CI/CD pipelines, and GenAI automation. From serverless workflows and microservices orchestration to GPU-enabled LLM clusters and RAG/agent pipelines, I deliver resilient, observable, and cost-efficient systems.
Expertise:
• Cloud-native architecture (AWS, Kubernetes, ECS, Cloud Run)
• CI/CD design, IaC (Terraform, Helm, Argo CD)
• AI infrastructure — LLMs, RAG, agents, PEFT, governance
• Automation at scale — n8n, Bedrock, LiteLLM, custom MCPs
Let’s make your infrastructure smarter, faster, and autonomous.
🔗 linkedin.com/in/nestorcolt
English
Native or bilingual
Spanish
Native or bilingual
Remote only
Primarily works remotely
Experience
- Leboncoin - Spain sourcing
On Malt
Software Engineer for GenAI-powered Internal Chatbot SRE / DevOpsTECHMarch 2025 - December 2025 (9 months)Onyx (Enterprise Gen AI Platform) – Deployed the full platform to production on Kubernetes with multi-node workers, scaling, and operational runbooks. Led enterprise onboarding with documentation and tutorials.• LiteLLM gateway – Centralized LLM access through automated key provisioning. Designed the workflow that generates virtual keys per team from a request form with usage tracking and quotas.• AWS Bedrock governance – Implemented inference profiles and tagging by team and business unit for cost tracking and accountability. Established basic guardrails and model routing policies.• ChatGPT Enterprise MCPs and internal GPTs – Created custom connectors for Jira, Confluence, and Backstage. Built internal GPTs exposing REST APIs for integration with internal systems.• n8n (company-wide LLM automations) – Deployed a multi-node Kubernetes cluster for large-scale workflow execution. Authored starter templates, documentation, and onboarding materials to drive adoption.• AKS GenAI enablement – Enabled GPU-based workloads across clusters and accounts to support LLM/ML inference. Standardized Helm deployments for LLM services and jobs.• Backups and DR – Implemented Velero for cluster and namespace backups with documented recovery procedures and tested restores.• Observability – Built Datadog dashboards for LLM gateways, GPU utilization, and performance metrics.• AI governance – Established practical processes for request handling, access issuance, tagging, and usage visibility across AI platforms. - The Cloud ClockWorkSolutions Architect | AI PlatformTECHNovember 2024 - Today (1 year and 7 months)Barcelona, Spain• Onyx (Enterprise Gen AI Platform) – Deployed the full platform to production on Kubernetes with multi-node workers, scaling, and operational runbooks. Led enterprise onboarding with documentation and tutorials.• LiteLLM gateway – Centralized LLM access through automated key provisioning. Designed the workflow that generates virtual keys per team from a request form with usage tracking and quotas.• AWS Bedrock governance – Implemented inference profiles and tagging by team and business unit for cost tracking and accountability. Established basic guardrails and model routing policies.• ChatGPT Enterprise MCPs and internal GPTs – Created custom connectors for Jira, Confluence, and Backstage. Built internal GPTs exposing REST APIs for integration with internal systems.• n8n (company-wide LLM automations) – Deployed a multi-node Kubernetes cluster for large-scale workflow execution. Authored starter templates, documentation, and onboarding materials to drive adoption.• AKS GenAI enablement – Enabled GPU-based workloads across clusters and accounts to support LLM/ML inference. Standardized Helm deployments for LLM services and jobs.• Backups and DR – Implemented Velero for cluster and namespace backups with documented recovery procedures and tested restores.• Observability – Built Datadog dashboards for LLM gateways, GPU utilization, and performance metrics.• AI governance – Established practical processes for request handling, access issuance, tagging, and usage visibility across AI platforms.• Cost optimization – Introduced tag-based showback via Bedrock profiles and LiteLLM usage metrics; provided guidance on right-sizing and off-peak scheduling.
- ClarivateCloud ArchitectPHARMACEUTICALS INDUSTRYJune 2022 - January 2025 (2 years and 7 months)Barcelona, Spain- Serverless event-driven pipelines for infrastructure management +240 AWS accounts- CICD end-to-end for Golde AMI solutions, architecture, and implementation- IAM lifecycle for all the accounts in the ORG- Serverless solutions for OpenSearch aggregators improving a complex company-wide logging platform- Sustainability and cost optimization
Recommendations
These freelancer profiles also match your criteria
Agatha Frydrych
Backend Java Software Engineer
4.7
(3)
2
Baptiste Duhen
Fullstack developer
4.6
(4)
5
Amed Hamou
Senior Lead Developer
4
(2)
7
Audrey Champion
Web developer
4.3
(3)
4
Education
- AWS Solutions ArchitectAWS2024Cloud Architect