SMASH, Who we are?
We are agents for tech professionals across Latin America who help them build careers with companies in the United States.
We believe in long-lasting relationships with our talent. We invest time getting to know them and understanding what they seek as their professional next step.
We aim to find the perfect match. As agents, we pair our talent with our US clients, not only by their technical skills but also by cultural fit. Our core competency is finding the right talent fast.
We purposefully move away from the traditional “outsourcing” relationship. Our clients are looking for professionals who become an integral part of their teams, and so are we.
Our Benefits
- Wellness Coverage
- Remote Work
- Birthday day off
- Recognition and rewards system
- Referrals Program
- Business skill coaching
- English classes for Smashers and relatives
- Learning opportunities
This is a remote position supporting a US-based company. Applicants must be legally authorized to work in their country of residence within Latin America (LATAM).
Role summary
We are looking for a Senior AI Automation Engineer to design and implement AI-driven solutions that automate complex lending, underwriting, document-processing, and operational workflows.
This role requires someone who can go beyond simply integrating an LLM. The ideal candidate can determine which AI models and architecture should be used for different use cases, connect AI systems with existing software and data sources, process structured and unstructured information, and build reliable automation around business-critical financial workflows.
Key use cases include document intelligence, workflow automation, automated loan-file review, underwriting and pricing rules, data analysis, email and messaging automation, and development of an intelligent loan-pricing solution.
Responsibilities
- Design and implement AI-powered automation solutions for lending and operational workflows.
- Evaluate different AI/LLM models, tools, and architectures and determine the appropriate technology for each use case.
- Design AI solutions considering cloud, on-premises, and hybrid deployment options.
- Integrate AI capabilities with existing proprietary and licensed software platforms, databases, websites, and internal applications.
- Build solutions capable of interpreting and analyzing structured and unstructured information from:
- PDF documents
- Excel spreadsheets
- Word documents
- Scanned documents and images
- Emails and attachments
- Phone and SMS logs
- Websites and web applications
- Databases and enterprise applications
- Develop document intelligence and data extraction workflows, including validation and interpretation of extracted information.
- Automate operational tasks and multi-step business workflows.
- Develop AI-powered processes to compile and present information for data-informed decision-making.
- Automate form completion and data entry processes.
- Build workflows capable of reading, generating, and processing emails and text-based communications.
- Generate automated reports, summaries, and data analyses.
- Translate existing underwriting and pricing rules into repeatable automated workflows.
- Apply standardized rubrics to evaluate properties, borrowers, and loan scenarios.
- Identify errors, missing information, inconsistencies, and potential non-compliance within processing requests.
- Develop an automated final loan-file review process that validates 20+ required data points before funding.
- Implement controls to verify both the completeness and accuracy of critical loan information.
- Design appropriate validation, auditability, and human-review mechanisms for AI-generated results.
- Work with business stakeholders to understand lending workflows and convert them into technical requirements and automated solutions.
Loan Pricing Automation
A major initiative for this role will be designing and implementing an intelligent loan pricing solution capable of:
- Collecting the required loan scenario information from borrowers.
- Validating borrower-submitted information before pricing.
- Processing borrower and property characteristics against lending criteria.
- Evaluating loan scenarios against multiple lender pricing sources and eligibility criteria.
- Extracting pricing information from PDF and Excel lender rate/pricing sheets.
- Accessing available online lender pricing tools and websites, including authenticated sources where technically and contractually permitted.
- Normalizing information received from different lenders and formats.
- Applying lender-specific rules and pricing logic.
- Generating multiple relevant loan options based on the borrower’s scenario.
- Presenting loan options in a structured and understandable format.
- Maintaining traceability between source information, pricing logic, and generated results.
- Supporting updates as lender criteria, pricing sheets, and underwriting requirements change.
Requirements – Must-haves
- Strong professional experience building AI/LLM-powered applications or automation solutions.
- Hands-on experience integrating LLMs and AI models through APIs into production applications or workflows.
- Strong software engineering and programming skills, preferably with Python.
- Experience designing AI agents, automated workflows, or multi-step AI systems.
- Experience processing structured and unstructured documents programmatically.
- Experience with document extraction, OCR/document intelligence, and data validation workflows.
- Experience working with PDF, Excel, Word, image, and other business-document formats.
- Experience integrating systems through REST APIs, databases, webhooks, and other integration mechanisms.
- Strong understanding of databases and structured data processing.
- Experience designing integrations with existing enterprise or proprietary software.
- Experience implementing workflow and process automation.
- Ability to evaluate and select appropriate AI models, architecture, infrastructure, and tooling based on business and technical requirements.
- Understanding of cloud-based and on-premises AI deployment considerations.
- Experience implementing validation and quality controls around AI-generated outputs.
- Strong understanding of data analysis and data transformation.
- Ability to translate complex business processes and decision rules into technical solutions.
- Strong problem-solving skills and ability to work with ambiguous or evolving requirements.
- Strong communication skills and ability to work directly with business stakeholders.
Nice-to-haves (optional)
- Experience within mortgage lending, commercial lending, financial services, fintech, or underwriting.
- Experience automating loan origination, underwriting, pricing, or funding workflows.
- Familiarity with Loan Origination Systems (LOS) or similar lending platforms.
- Experience with RAG (Retrieval-Augmented Generation) architectures.
- Experience developing production-grade AI agents or agentic workflows.
- Experience with browser automation and authenticated web workflows.
- Experience integrating email and messaging platforms.
- Experience with cloud AI services on Azure, AWS, or GCP.
- Experience with vector databases and enterprise search technologies.
- Experience implementing human-in-the-loop AI workflows.
- Experience with AI evaluation, observability, prompt management, and model performance monitoring.
- Understanding of security, privacy, auditability, and access-control considerations for sensitive financial information.
- Experience building analytical or rules-based pricing engines.
Languages