About the Role

We are expanding the team behind a production-grade document intelligence platform that ingests purchase agreements and related contracts, processes them through OCR and large language models, and extracts contractual obligations, counterparties, and key terms with full clause-level traceability.

This is a highly autonomous, hands-on engineering role where you will own features end-to-end, from data modeling and backend services to AI integrations, UI implementation, and cloud infrastructure.

What You'll Do

- Extend the document extraction pipeline to support new agreement types and contract facts.
- Improve recall and extraction quality for complex, multi-party documents.
- Build verification and grounding mechanisms to validate model outputs and escalate uncertain results for human review.
- Develop classical ML solutions for document routing, confidence scoring, and entity matching.
- Enhance reviewer workflows used by legal teams and asset managers.
- Maintain and improve GCP infrastructure, deployment pipelines, and platform reliability.
- Collaborate with non-technical stakeholders to translate business requirements into system behavior.

Required Qualifications

- Strong Python development experience with Python 3.9+.
- Experience with Flask (or FastAPI/Django and willingness to work with Flask).
- SQLAlchemy 2.0 and PostgreSQL expertise.
- Strong relational database modeling skills.
- Production experience building LLM-based document extraction solutions.
- Experience with structured outputs and JSON schemas.




- Experience working with Vertex AI, Bedrock, Azure OpenAI, or similar platforms.
- Deep understanding of OCR workflows and PDF processing.
- Experience evaluating non-deterministic AI systems through testing, regression analysis, and recall measurement.
- Practical machine learning experience using scikit-learn or equivalent.
- Experience with Google Cloud Platform including Cloud Run, Cloud SQL, Cloud Storage, Secret Manager, Identity-Aware Proxy, and Vertex AI.
- Terraform, Docker, and CI/CD experience.
- Frontend experience with Jinja2, Tailwind CSS, vanilla JavaScript, accessibility standards, and semantic HTML.
- Strong testing discipline using pytest and JavaScript testing frameworks.

Preferred Qualifications

- Microsoft Graph API and SharePoint integration experience.
- Legal-tech, compliance, or contract management domain experience.
- Renewable energy or power markets exposure.
- Experience working directly with business stakeholders.

Technical Environment

- Python, Flask, Jinja2
- Vertex AI, Gemini, Claude, scikit-learn
- Google Document AI, pypdf
- PostgreSQL, SQLAlchemy 2.0
- GCP, Cloud Run, Cloud SQL, Cloud Storage
- Docker, Terraform, GitLab CI
- Tailwind CSS v4, Vanilla JavaScript
- pytest, jsdom

Why Join Us

- Work on a real-world AI product in production.
- Build systems that combine LLMs, OCR, ML, backend engineering, and cloud infrastructure.
- Collaborate closely with legal and asset-management stakeholders.
- Own high-impact features from concept to deployment.

📌 Senior AI / Full-Stack Engineer — Document Intelligence (Colombia)
🏢 Unosquare
📍 Colombia