Forward-deployed AI engineer, Developer Tools → Belo Horizonte · Recife · Granada · San Francisco ·

Joinplank; Belo Horizonte, Recife; On-site; BRL 180,000 - 300,000; Full time

Job summary

Joinplank is seeking an experienced software engineer to embed with customer teams on-site, owning delivery end to end from discovery to production rollout. You will write and review production-grade code across the stack, with a focus on Python and strong familiarity with TypeScript.

The role involves deploying LLM-powered solutions at scale, building agent infrastructure, and collaborating closely with customers and leadership.

Qualifications

• 4–10 years of engineering experience, customer-facing or deployment preferred.
• Proficient in Python; TypeScript is a strong plus.
• Hands-on production experience with LLMs, prompts, evals, and deployment at scale.
• Experience with cloud, data pipelines, and enterprise environments.
• Strong communication and stakeholder management skills.

Responsibilities

• Embed with a customer team, on-site, owning delivery end to end.
• Drive 0-to-1 prototypes to MVP to scale, shipping production-grade code.
• Build agents, sub-agents, MCP servers, and RAG pipelines.
• Be the primary technical owner of the engagement with customer engineers.
• Codify patterns into reusable blocks and playbooks.
• Surface deployment friction and provide eval-driven signals to inform product.
• Build agent infrastructure, SDKs, and developer-facing APIs.

Skills

Python

TypeScript

LLMs

Cloud

Data pipelines

Stakeholder management

Tools

APIs

Agent frameworks

Job description

Plank places forward-deployed engineers inside developer-tools teams to take AI into production in the infrastructure other builders depend on. Plank already builds agent infrastructure at Near AI, led by a co-author of Attention Is All You Need, and AI code review at Entelligence. You own delivery end to end, vendor-neutral across OpenAI, Anthropic, and open models.

What you will do

• Embed with a customer team, often on-site, and own delivery end to end: discovery, technical scoping, system design, build, and production rollout.
• Drive solutions from a 0 to 1 prototype through MVP to scale, writing and reviewing production-grade code across the stack.
• Build the hard parts hands-on: agents and sub-agents, MCP servers, RAG pipelines, evals, and integrations against real systems of record.
• Be the primary technical owner of the engagement, partnering with the customer's engineers, domain experts, and leadership.
• Codify what works into reusable building blocks, playbooks, and repeatable deployment patterns.
• Bring the field back to product: surface deployment friction and eval-driven signal that sharpens what gets built next.
• Build agent infrastructure, SDKs, and developer-facing systems, and design APIs other engineers will build on.

What we are looking for

• 4 to 10 years of engineering experience, ideally customer-facing or in deployment; former founders welcome.
• You write and review production-grade code across frontend and backend, with Python as a baseline and TypeScript a strong plus.
• Hands-on production experience with LLMs: prompt engineering, agent development, evals, RAG, and deployment at scale.
• A track record of scoping and shipping complex systems, and comfort with cloud, data pipelines, and enterprise environments.
• High agency and comfort with ambiguity: you turn vague business problems into concrete technical plans and own the outcome.
• Strong communication and stakeholder management, from technical discovery to the executive relationship.
• Interest or experience in developer tools, infrastructure, or platform engineering.

Nice to have

• Open-source work, or experience building SDKs, APIs, or agent frameworks.