Required Skills & Experience
Must-Have Experience (in priority order)
- Systems where the number has to be right. You have built billing engines, payment settlement, accounting, reconciliation, claims or trading back-office systems, and you know what happens when a system gets a number wrong.
- Rules that change over time. You have built systems that apply the rule in force on each transaction's date, for example billing with changing price plans or insurance with policy versions.
- Corrections and reversals. You have handled refunds, cancellations or adjustments that unwind prior results and trigger recalculation.
- Explainable results. You have built audit trails that tie every output to its inputs, rule version and evidence.
- End-to-end ownership. You have personally taken a production system from ambiguous requirements through design, build and operation. Experience & Technical Requirements
- 10+ years of professional software engineering experience, with at least 3 years in a staff, principal or lead architect role.
- 5+ years building financially sensitive systems such as billing, payments, settlement, accounting, reconciliation, claims or trading back-office.
- 5+ years of backend development in Python, Java, C#, Go or a similar language. Python is preferred.
- 5+ years of SQL and relational data modeling (PostgreSQL, SQL Server, Oracle or similar), including: ○ query tuning on datasets with millions of rows ○ effective-dated or temporal data designs
- 3+ years with containers and CI/CD: Docker, Kubernetes or similar, plus GitHub Actions, GitLab CI, Azure DevOps or Jenkins.
- 2+ deployments into customer-controlled or on-premises environments, including restricted-network or offline setups.
- Experience with at least one major cloud platform (AWS, Azure or GCP) and infrastructure-as-code (Terraform or similar).
- Hands-on use of AI coding agents (such as Claude Code, Cursor, GitHub Copilot or Codex) in real production delivery, not just experimentation. Strongly Preferred
- On-premises delivery: Deployed inside a customer's own environment, including offline and with their security team involved. Cloud-only experience is a gap for this role.
- AI-native engineering: Directs coding agents as a core part of delivery, and reviews, tests and owns their output.
- High-volume data: Backend, SQL and data modeling for systems that process millions of rows per run.
Nice to Have Skills & Experience
- Domain background: Loyalty, rewards or card programs. Any rules-heavy financial domain is equally valuable.
- Compliance frameworks: SOC 2, ISO 27001 or similar.
- Preferred technologies: ○ rules engines or domain-specific languages (DSLs) ○ event sourcing and temporal tables ○ Kafka or similar streaming ○ Airflow, dbt or Spark ○ secrets management (Vault, Azure Key Vault, AWS KMS) document extraction with source-level provenance
Job Description
Our client is a fintech company in the loyalty and payments space. They are building a platform that checks transactions against contract terms and program rules, flags discrepancies, and backs every finding with evidence. They need a hands-on technical owner to design, build and run the first release. There is no one above this person technically: you will own the architecture, the code and the outcome. The core challenge is building a calculation system that is deterministic and provably correct. Every result must apply the rules in force on the transaction date, and it must survive later corrections and reversals. It also has to trace back to the exact inputs and evidence that produced it. This is not a data science or machine learning role. The system must never guess. What You'll Do
- Design and build a rules engine that evaluates transactions against versioned, effective-dated rules, including: ○ precedence and exceptions ○ caps and tiers ○ balances and rounding
- Handle corrections and reversals. When a refund or cancellation unwinds earlier results, everything downstream must be recalculated predictably, with the original results kept.
- Make every figure explainable, traced to its inputs, rule version and supporting evidence through a real audit trail rather than logging.
- Turn ambiguous business requirements into clear designs, data models and acceptance criteria, working directly with domain experts.
- Deploy into customer-controlled and on-premises environments, including offline setups and alongside customer security teams, as well as cloud.
- Use AI coding agents as a core part of delivery, while keeping full accountability for what they produce.
- Choose the stack. Python is likely, but the first engineer decides. What Success Looks Like
- Business rules are documented before they become hidden assumptions in the code.
- The first release produces accurate, reproducible findings backed by clear evidence.
- Late corrections and rule changes are handled predictably and explicitly tested.
- The system deploys securely in both cloud and customer environments.
- Another engineer can understand, run and extend the system from its documentation and tests.