Remote Forward-Deployed AI Engineer (Python)

FullStack; Perímetro Urbano Barranquilla; Remote; COP 384,813,000 - 577,219,000; Full time

Benefits offered by this job

100% remote work

Health policy with Sur a

English classes

Food card

Career advancement

Job summary

FullStack is seeking Forward Deployed Engineers who embed with enterprise teams to architect, build, and sustain AI-powered solutions. You will balance consultant, operator, and engineer roles, tackling data, retrieval, and model behavior to ensure trustworthy agents while aligning with client business goals.

You will work with regulated industries and Fortune 500 clients, leading discovery discussions and delivering working prototypes within weeks, not months.

Qualifications

• 5+ years of professional software development experience with at least one mainstream production stack.
• Experience shipping agentic AI solutions in production environments.
• Advanced English is required.

Responsibilities

• Discover and run workshops with business, data, and engineering stakeholders.
• Structure process and specifications for automation.
• Architect end-to-end agentic systems with retrieval and context.
• Build working systems: agents, retrieval services, and eval tooling.
• Operate Day 2: monitor valuation, observability, drift, and handoff to client teams.
• Consult with CTOs, VPs, and business leadership on decisions and impact.

Skills

Python

TypeScript/Node.js

Java

Go

C#/.NET

Agentic AI

LLM workflows

RAG architectures

Education

Four-year college degree

Tools

Cursor

Claude Code

Gemini CLI

OpenAI Codex

Job description

FullStack is seeking Forward Deployed Engineers who embed with enterprise teams to architect, build, and sustain AI-powered solutions. You will balance consultant, operator, and engineer roles, tackling data, retrieval, and model behavior to ensure trustworthy agents while aligning with client business goals.

You will work with regulated industries and Fortune 500 clients, leading discovery discussions and delivering working prototypes within weeks, not months.