AI Senior Engineer (Vision)

Role overview A 100% remote role within Latin America seeks a senior engineer to own the intersection of computer vision and language model orchestration. The work involves extracting structured meaning from complex visual documents and wiring that data into LLM-driven reasoning flows. It suits someone who enjoys both the perceptual layer, the “eyes” of the system, and the agentic logic that decides what those eyes mean.

Responsibilities - Build pipelines that parse complex documents, understanding layout, charts, and visual context through vision-language models and dedicated layout analysis. - Implement orchestration layers with LangChain or LangGraph to drive agents, tool calls, and chain-of-thought workflows. - Process native PDFs using libraries such as PyMuPDF, preserving structural fidelity before model ingestion. - Craft prompts and control flows that keep models accurate on financial charts and dense layouts while minimizing hallucination. - Apply cost-aware engineering choices, including batching and model selection, so vision and orchestration remain economically sustainable.

Requirements - Deep experience with LLM orchestration frameworks (LangChain, LangGraph, or comparable), including context window management and agentic workflows. - Hands-on integration of frontier vision models such as GPT-4V and Claude 3.5 Sonnet, plus embedding models like CLIP. - Familiarity with document-intelligence tooling such as Donut, Pix2Struct, Unstructured.io, or Docling. - Mastery of PDF extraction libraries such as PyMuPDF or pdfplumber. - Strong Python machine learning background with PyTorch or TensorFlow. - Excellent verbal and written English communication.

Nice to have - Fine-tuning experience on vision or language models for domain-specific artifacts such as financial tables and charts. - Prior exposure to real estate or financial document workflows.

Benefits and work setup - Fully remote across LatAm. - 40 hours per week during regular business hours. - Compensation paid in USD. - 18 days of paid time off annually, plus observance of local holidays and a winter break between Christmas and New Year. - Wellness and remote stipends and an AI learning voucher.