Skip to content
Halyard Commercial SystemsContact us
Menu

Build with Halyard

Full Stack Software Engineer

AI systems & integrations · Financial services

Build AI systems that financial services teams can depend on.

Take a real operational problem from the first conversation through to a working system. You’ll help design the data behind it, connect the existing platforms, build the interfaces people use and make the whole thing reliable enough for everyday use.

Apply for this role

Starting compensation

USD $50,000–$75,000 per year, depending on experience.

Engagement

Initially an independent contractor engagement.

Location

Fully remote. North or South America preferred, particularly the US or Canada. Applicants based in Latin America are welcome.

Working hours

Aligned with US Eastern time. You do not need to live in the Eastern time zone.

The opportunity

A chance to shape more than the next feature.

Halyard builds AI-enabled commercial systems for financial services firms. We connect their existing platforms, data and institutional knowledge so their teams can serve more clients and manage growing demands without every increase in volume creating more manual administration.

Our focus includes fund administrators, investment managers, investment funds, corporate services providers and trust and fiduciary firms. These businesses depend on accurate records, experienced people and clear accountability. Useful AI needs to fit that environment.

We’re looking for developers who want meaningful ownership of complete capabilities, from a sound data model to a usable interface and a dependable deployment. Our ambition is to build systems that become a lasting part of how financial services firms operate. Joining Halyard at this stage means contributing to the engineering patterns and reusable foundations behind future client solutions.

You’ll collaborate with Halyard’s commercial and engineering team and, where appropriate, client teams and other technology providers. Ownership means being accountable for your agreed scope, communicating trade-offs and asking for the right input. It doesn’t mean being expected to solve every architecture, business or regulatory question alone.

The product direction

What you could help build

Our product direction is a connected operating layer around the platforms a firm already uses. A scoped implementation might include capabilities like these:

Reporting preparation and review

Bring approved system exports, supporting documents and reporting instructions into a reviewable package. Give staff a clear view of missing inputs, queries, versions and approvals while the existing accounting platform remains authoritative for its calculations.

Client and entity information

Connect client, fund, investor and legal-entity records without confusing a person’s roles across different relationships. Reuse confirmed information in approved documents and show what still needs checking before a record is updated.

Operational knowledge and coordination

Make permitted records and relevant prior decisions searchable with links back to the source. Connect that context to tasks, owners and review queues so someone can understand what is outstanding and take the next authorised step.

These examples illustrate the kinds of systems we aim to build. Each client engagement has its own agreed scope, technical constraints and approval requirements.

The role

What you’ll be responsible for

Turn business requirements into a buildable design

Work with the team to understand who uses a process, what information it depends on and which decisions belong to people or existing systems. Break the agreed scope into interfaces, data contracts, deliverables and testable acceptance criteria. Raise missing requirements before they become implementation problems.

Design the shared data model

Model clients, entities, funds, documents, relationships and process states with clear identifiers and ownership. Handle versions, conflicting inputs and data quality deliberately. Make it clear which system owns a record and how an approved change is reflected elsewhere.

Build reliable connections between systems

Develop APIs, webhooks, background jobs and controlled file imports or exports where appropriate. Handle authentication, validation, rate limits and reconciliation. An interrupted request or repeated event should have a defined recovery path rather than leave the team guessing what happened.

Make multi-step processes dependable

Build persistent workflow state, task ownership, queues, approvals and exception routes. Preserve progress across failures and restarts, and prevent duplicate actions. Tie approvals to the relevant version of the information so a later edit cannot silently inherit an earlier sign-off.

Integrate AI with clear boundaries

Use LLMs and retrieval where they help interpret requests, extract information or prepare drafts. Validate structured outputs, retain source references and evaluate against representative cases. Keep permissions, business rules and consequential actions under explicit application controls and the required human review.

Build interfaces people can use with confidence

Develop responsive internal tools and, where needed, client-facing interfaces. Make sources, changes, outstanding questions and approval status easy to inspect. Support correction and exception handling as carefully as the ordinary path through the system.

Ship, monitor and leave a maintainable system

Write automated tests, review code and support controlled releases. Build in access controls, client separation, secure secret handling and useful audit records. Document deployment, monitoring, recovery and configuration so another authorised engineer can understand and maintain the system without depending on you to remember every detail.

Your background

The experience we’re looking for

You should have enough production experience to take a defined capability through design, implementation, testing and release, with appropriate technical review.

  • Full stack applications you’ve helped put into production, with a clear explanation of the parts you personally designed, built and maintained.
  • Strong JavaScript or TypeScript skills, a modern frontend framework and practical backend development in Node.js, Python or a comparable environment.
  • Solid relational data modelling and SQL, including migrations, constraints and the relationships between records.
  • Experience integrating external systems and handling authentication, background processing, partial failures and data reconciliation.
  • Practical testing, debugging, Git-based collaboration and cloud deployment skills, with an understanding of server-side authorisation and protecting sensitive data.
  • Clear communication. You can explain a technical choice to a colleague, surface uncertainty early and keep an agreed delivery scope moving.

Especially useful, but not a checklist you must complete

  • Production LLM integrations, document processing, retrieval, evaluation or AI-assisted development.
  • Workflow engines, event-driven systems, audit trails or applications with complex approval and access rules.
  • Financial services, accounting, entity management or another environment where data accuracy and traceability matter.

You don’t need to arrive knowing every financial services platform. You do need to be able to learn an unfamiliar system, ask precise questions and distinguish what you’ve verified from what you’re assuming. We care more about sound engineering judgement and evidence of delivery than a long list of framework names.

Our standard

How we approach engineering

Understand the business consequence

A technically impressive feature still has to make a real process easier to run. Understand what the user needs to achieve and what a mistake would mean before choosing the solution.

Be honest about what is ready

A useful demonstration is a step forward. Production readiness also needs evidence from testing, review and operation. Be direct about what is complete, what is uncertain and what needs help.

Use AI thoughtfully

Use AI tools to accelerate development where appropriate, within agreed confidentiality and security boundaries. You remain responsible for understanding, testing and explaining the code and behaviour you ship.

Build for the next engineer and the end user

Prefer a clear, proportionate design that the team can operate and extend. Documentation, recovery and handover are part of delivering a useful system, not tasks to leave until someone asks.

What good looks like

For an agreed first capability, you can explain the data model and integration boundaries, ship a tested path through the process and show what happens when inputs are incomplete or a connected service fails. The user can review the relevant evidence, correct a draft and see who owns the next decision.

The wider team can inspect the code, deploy it through the agreed process and understand its operating requirements. As we add capabilities, the foundations become easier to reuse without mixing client data or carrying one client’s rules into another client’s system.

Apply to Halyard

Show us something you’ve built.

Tell us about a relevant system, the part you built and one decision you made. A short explanation is enough, even when the code is private.

Please only share material you’re authorised to disclose. You don’t need to include confidential client information.

Our team will review your application and reach out if we’d like to schedule an interview.

For example America/Toronto or America/Sao_Paulo
Use a complete https:// link
What did it need to do, what did you personally build and how did you handle an engineering decision or something that went wrong?
Include the frontend, backend, database and integration technologies most relevant to this role.
Let us know about any notice period or commitments that affect your availability.
PDF or DOCX, up to 2 MB