Esteban Gaviria

Hi, I'm Esteban Gaviria

AI Engineer

I build LLM, RAG and real-time voice systems, and ship them to production.

Cali, Colombia · Open to remote

AI Engineer CV DevOps Engineer CV
Portrait of Esteban Gaviria Zambrano

About

I'm an AI engineer from Cali, Colombia. At Brain AIX I build conversational AI agents that talk with a cloned voice and an avatar: the way they use language models, look up each user's documents and answer by voice in real time.

I also take care of getting that work to production, from containers to CI/CD. I'm finishing a double degree in Telematics and Systems Engineering at Universidad Icesi, and I compete in judo.

Currently
AI Software Developer at Brain AIX
Studying
Telematics & Systems Engineering, Universidad Icesi
Languages
Spanish (native), English (C1)

Experience

Brain AIX

AI Software Developer · Remote

September 2025 – Present

Conversational AI platform where agents talk in real time with a cloned voice and an avatar.

  • Built the LLM orchestration layer: memory, tool use and streaming across OpenAI, Anthropic, Gemini and open-source models.
  • Built the RAG pipeline that grounds each agent in its user's own documents, with a fine-tuned embedding model.
  • Built the real-time, multilingual voice pipeline, from speech recognition to voice-cloned speech.
How a conversation flows Simplified view of the voice and retrieval pipeline. Provider names are the ones used in production; internal details are omitted.
Voice
  1. User speaks
  2. Speech-to-text Deepgram · Whisper
  3. LLM orchestration memory · tools · streaming
  4. Text-to-speech voice cloning
  5. Avatar answers
Grounding
  1. User documents PDF · Word
  2. Chunking
  3. Embeddings fine-tuned model
  4. Vector search context for the LLM

Universidad Icesi

2023 – 2026

Teaching Assistant · Mathematics & Computational Thinking

Supported engineering students in applied mathematics, calculus and programming fundamentals.

Projects

MicroTodoSuite

Infrastructure Lead · team of 3

April 2025 – Present

A microservices platform my team of three moved from Azure to Kubernetes on AWS, with dev, staging and production environments.

  • Built AI agents and skills that write, review and audit the project's infrastructure code.
  • Led the migration to Amazon EKS, with all the infrastructure written in Terraform.
  • Designed an economical and a full architecture, plus a recovery environment on Azure with tested failover.
From a merged pull request to a running cluster The delivery path every service shares. Nothing reaches a cluster by hand.
Delivery
  1. Service repo thin CI caller
  2. Reusable CI build · Trivy · SBOM · Cosign
  3. Amazon ECR signed image
  4. GitOps PR digest bump
  5. ArgoCD Kyverno verifies
  6. Amazon EKS dev · staging · prod
Foundations
  1. Terraform roots saved plans only
  2. AWS VPC · EKS · IAM · Route 53
  3. Azure AKS warm standby

Skills

LLM engineering

  • LLM orchestration
  • Prompt engineering & evaluation
  • RAG
  • Embeddings & fine-tuning
  • Weaviate
  • Agents & tool use (MCP)
  • LangChain
  • LangGraph
  • LlamaIndex
  • Hugging Face

Models & voice AI

  • OpenAI
  • Anthropic Claude
  • Google Gemini
  • Alibaba Qwen
  • Open-source LLMs
  • Deepgram
  • Whisper
  • ElevenLabs
  • Azure/Google Speech
  • Voice cloning

Software engineering

  • Python
  • FastAPI
  • TypeScript
  • Node.js
  • NestJS
  • React
  • Next.js
  • Java
  • Bash
  • PostgreSQL
  • MongoDB

AI-assisted engineering

  • Claude Code
  • Codex
  • Cursor
  • Antigravity
  • Custom skills
  • MCP servers

Cloud

  • AWS EKS
  • EC2
  • VPC
  • IAM/IRSA
  • ECR
  • Route 53
  • Transit Gateway
  • Azure AKS
  • Azure Container Apps
  • GCP GKE

Kubernetes & delivery

  • Kubernetes
  • Docker
  • Kustomize
  • Istio
  • Argo Rollouts
  • ArgoCD
  • GitHub Actions
  • Terraform
  • tflint
  • Jenkins

Security & observability

  • Trivy
  • Syft
  • Cosign
  • Kyverno
  • External Secrets
  • Prometheus
  • Grafana
  • Loki
  • Jaeger
  • Infracost

Education & credentials

Education

Universidad Icesi · Cali, Colombia

B.Sc. Telematics Engineering & B.Sc. Systems (Software) Engineering, double degree

Final semester · expected graduation August 2027

Certifications

  • EF SET English Certificate

    C1 Advanced, 2025

  • CNCF Kubestronaut path

    CKA scheduled for January 2027; KCNA, KCSA, CKAD and CKS for March 2027

Beyond the code

  • Community

    CNCF Colombia member since 2025; Kubernetes Community Days Bogotá 2025 attendee.

  • Judo

    Bronze medal (3rd place), −90 kg, ASCUN 2026 National University Games, representing Universidad Icesi.

Contact

Open to AI engineering and DevOps roles, remote or hybrid. The fastest way to reach me is email.

Email me

estebangaviria82@gmail.com

Or send a message