Oklahoma · available for remote roles

Senior Java backend developer for complex production systems.

7+ years building and maintaining backend settlement platforms, data-processing workflows, and calculation-heavy enterprise systems in the energy sector.

View Engineering Work → Download Resume GitHub LinkedIn

7+ yrs

Backend engineering

CAISO

Lead settlement developer

3 ISOs

CAISO · PJM · ERCOT

Engineering Dossier

REF · KR-2026

System overview

{ }

ISO Data

✓

Validation

Java Svcs

Database

≡

Reports

Contents

Production Debugging§03

Regulatory Change Implementation§03

JUnit Regression Coverage§04

Oracle Data Validation§02

Overview

Backend Systems

Design, build, and maintain the Java/Spring services that process market data and compute settlements — the layer where correctness is enforced.

Production Reliability

Keep regulated daily and monthly settlement cycles on schedule, debugging live issues and preventing regressions before they reach clients.

Settlement Domain

Deep working knowledge of ISO market operations — CAISO, PJM, ERCOT — and how tariffs and charge codes translate into calculations.

Java / SQL / Testing

Java 8/11, Spring Boot, Oracle SQL, and a testing discipline built on JUnit, Mockito, and regression coverage around financial logic.

How the systems fit together

A settlement platform end to end. Select a stage to see where I own the work.

STAGE 03

Java Services

spring boot · charge-code logic

Engineering work

CAISO · charge codes

Regulatory Calculation Implementation

Turning CAISO charge-code requirements into Java logic, SQL mappings, and test coverage — correct on the regulatory effective date.

Problem

A CAISO tariff change introduced revised charge codes with a fixed regulatory effective date and no grace period for miscalculation.

Constraints

Hard deadline, backward compatibility with historical runs, and full auditability back to tariff language on financially material charges.

Technical Work

Modeled the charge-code logic in Java/Spring directly against the tariff, mapped inputs through SQL, and made effective dates configurable so the switchover left prior periods untouched.

Validation

Built JUnit cases from tariff worked examples, verified outputs with business analysts, and ran regression across historical settlements to confirm no drift.

GhgAllocationCharge.java

@Service
class GhgAllocationCharge {
  BigDecimal compute(SettlementInterval iv) {
    return iv.allocatedMwh()
      .multiply(tariff.ghgRate(iv.date()))
      .setScale(2, HALF_UP);
  }
}

Result

Charge codes went live on the effective date with values matching expected results — no post-go-live corrections.

production · debugging

Production Issue Debugging

Tracing settlement discrepancies across data inputs, calculation logic, database state, and reports — under a live delivery deadline.

Problem

A monthly settlement run surfaced discrepancies between calculated charges and expected values, putting an on-time client deliverable at risk.

Constraints

A live production cycle with a hard window, large Oracle datasets, and an unknown root cause spanning data, logic, and timing.

Technical Work

Traced affected charges from report output back through the calculation logic and validation layer to source intervals, isolated a data-input anomaly, and coordinated the fix with operations.

Validation

Reconciled reprocessed results against expected values and prior periods, then added an automated check to catch the same class of anomaly in future runs.

Result

Discrepancy resolved and the deliverable met its window; a new guardrail now prevents recurrence.

CAISO · PJM · ERCOT

ISO Data Processing

Working across file-based ETL, XML/JSON, Oracle queries, and settlement reporting for multiple market structures.

Problem

Settlement inputs arrive from multiple ISOs, each with its own file formats, market structure, and timing.

Constraints

Heterogeneous formats and schedules, large datasets extracted from Oracle, and required accuracy across every market.

Technical Work

Built Java processing and SQL/stored-procedure ETL to normalize market-data files into a consistent internal model, with per-ISO handling isolated behind a shared interface.

Validation

Automated validation tools verified record integrity and inputs per ISO, checked against source files and expected settlement outputs.

Result

Reliable, repeatable multi-ISO processing feeding accurate settlement runs across distinct markets.

Technical stack

Backend

Java 8/11 · Spring Boot · Hibernate · JPA · Maven

Data

Oracle · SQL Server · PostgreSQL · SQL validation · stored procedures

Testing

JUnit · Mockito · regression testing · code reviews

Domain

CAISO · PJM · ERCOT · ISO settlements · market operations

How I work

The habits that keep financially material systems correct and clients unbothered.

01Implements backend logic directly from regulatory requirements.

02Debugs production settlement issues down to the source interval.

03Writes regression tests around financial calculations.

04Maintains legacy systems without breaking client workflows.

05Communicates clearly with analysts, operations, and client teams.

06 · CONTACT

Let's talk about the systems you need to keep running.

Open to remote senior backend and Java roles in fintech, energy, trading, and enterprise platform teams.

Emailhello@kyleroller.dev ↗ LinkedInlinkedin.com/in/kyleroller ↗ GitHubgithub.com/kyleroller ↗ ResumeDownload PDF ↗