Forward Deployed Engineer (Contract) - LATAM

Contract, project-based · Remote

The role:

Dual Logic helps organizations of 50 to 1,000 people put generative AI to work. We're looking for

experienced technologists to work with us on client projects as Forward Deployed Engineers. You'll work directly with client teams. You'll find the real problem, scope the solution, and implement.

Much of our work doesn't start from a blank page. You'll take over prototypes that staff built themselves, codebases handed over without documentation, and agents that only one person understands. Your job is to make them safe, reliable, and adopted. When no existing tool fits, you'll build a new one.

The work comes one project at a time. Before you agree to one, we'll tell you what it involves, when it runs, and how much time it takes. Some projects are short, fixed-scope reviews. Others are builds or retainers that run for several months. The time commitment ranges from a set number of hours a month to close to full time. You can take on one project or several, depending on your availability and fit.

Preference will be given to individuals that can take on more than one engagement or project at a time.

What you'll do

• Assess what exists. Understand the client’s technology stack and existing ecosystem of tools,

including those that staff may have built without engineering support.

• Take over and rebuild inherited systems. Get up to speed on an unfamiliar codebase, stabilize

it, and move it to a better architecture without breaking what users rely on. Where accuracy

matters, use parity tests (tests that confirm the new version produces the same results as the old

one) to prove it.

• Build AI applications and agents. Ship retrieval systems, multi-step agents, and other custom

tools and applications that plug into the client's existing systems.

• Recommend platforms and defend the choice. Weigh deployment options against the client's

actual environment, budget, and skills. Guide the client on selecting the right tool for the right job,

and explain the tradeoffs in plain language.

• Design secure integration layers. Build reusable services that connect AI tools to core business

systems. Use service accounts, per-solution keys, and secret management, and get the work

through the client's security review.

• Drive adoption, not just delivery. Work with end users to find why a tool isn't being used, and fix

it. Set up feedback loops and measure time saved and usage.

• Transfer ownership. Document decisions, pair with client staff, and remove single points of

failure, so the capability stays after the project ends.

You'll work with a Dual Logic partner / engagement lead. Clients will see you as part of the Dual Logic

team.

What you bring

• A record of building and shipping production software, including work done directly with

customers or business stakeholders.

• Strong skills in Python or TypeScript, and comfort working in the other.

• Experience building on LLM APIs: prompting, retrieval, tool use, and multi-step or multi-agent

workflows. You also know where they fail.

• Solid backend fundamentals: relational databases (PostgreSQL), API design, authentication, and

deployment pipelines.

• A record of taking over code you didn't write and improving it safely.

• Security judgment around AI systems: secrets handling, data classification, least-privilege access, and working with information security teams.

• The ability to explain a technical tradeoff to an executive in a few sentences, and to tell a client

when their preferred option is the wrong one, nicely.

• Comfort with ambiguity. Scope changes during our engagements, and you'll help reset it.

• The ability to manage your own time well across more than one project.

Nice to have

• Microsoft ecosystem experience: Copilot, Copilot Studio, Azure AI Foundry, and Word/PowerPoint

document generation.

• Firebase (Firestore, Cloud Functions) or Supabase-style Postgres platforms with row-level security.

• React and Next.js front-end work.

• Workflow orchestration tools such as n8n.

• Experience implementing Model Context Protocol (MCP) servers and experience designing

gateways that expose data safely to builders who aren't engineers.

• Familiarity with AI governance frameworks, such as the NIST AI Risk Management Framework.

• Consulting or independent client work.

How we work

• Meet clients where they are. Some clients have one AI prototype. Others have an established AI

team. You'll adjust to both.

• Build with the people a solution is intended to serve. AI supports the people doing the work. It

doesn't replace them. A tool nobody uses is a failed project.

• Tie work to measurable targets. Every engagement has a way to tell whether it worked.

• Treat responsible use as everyone's job. Governance and security are part of the build, not a

review step at the end.

Terms

• Engagement: Independent contractor, one project at a time.

• Time commitment: Varies by project. We agree on it before you start.

• Location and travel: Fully remote. No travel required.