Senior Software Engineer - AI Development Tooling & Data Architecture

Senior Software Engineer - AI Development Tooling & Data Architecture

Contract Type:

Contract

Location:

Westbrook - ME

Industry:

Information Technology

Contact Name:

Contact Email:


Contact Phone:


Date Published:

11-17-2026

Salary:

$55 - $60 Per Hour

Job ID:

40521

Senior Software Engineer – AI Development Tooling & Data Architecture

Overview

Build the systems that make AI-assisted software development more reliable—and help establish how complex clinical data can be understood, measured, linked, and responsibly used. This fully remote role offers two entry points: AI development tooling or data architecture . You do not need to be an expert in both areas; depth in one, combined with strong reasoning and curiosity about the other, is enough to succeed.

You’ll create durable engineering practices that turn requirements, architectural decisions, data contracts, and measurement definitions into tools and models teams can trust. The work emphasizes evidence, reproducibility, clear technical writing, and sound judgment over familiarity with a particular technology stack or quantity of code produced.

Required Skills

  • Strong production experience with Python .
  • Ability to learn unfamiliar technical or business domains, identify important questions, and revise conclusions when evidence changes.
  • Strong evidence-based reasoning, including the ability to distinguish:
  • Measurements from inferences
  • Proposals from decisions
  • Structural possibility from demonstrated capability
  • Ability to turn grammars, contracts, invariants, type systems, or measurement definitions into useful tools and models.
  • Clear technical writing skills, with experience creating design documents, assessments, specifications, or decision records.
  • Independent delivery and sound judgment, including recognizing when a design decision or formal approval is required.
  • Experience using AI coding assistants, along with the judgment to identify plausible but incorrect, incomplete, or architecturally damaging output.
  • For the AI development tooling track:
  • Experience with Python’s standard library and ast, or equivalent code-as-data and program-analysis work.
  • For the data architecture track:
  • Strong SQL skills and analytical reasoning.
  • Deep understanding of data grain, joins, denominators, and population comparability.
  • Evidence of substantial self-directed learning followed by practical application.

Nice-to-Have Skills

Experience in one or more of the following areas is beneficial:

  • Static analysis, linters, codemods, or architecture-rule tooling
  • Schema-driven code generation
  • Property-based, golden, mutation, or contract testing
  • Model-graded evaluation, probe design, or measurement of nondeterministic systems
  • Data modeling, data catalogs, lineage, record linkage, dbt, Spark, or schema languages
  • Declarative rules, domain-specific languages, constraint systems, or information theory
  • Reproducible tooling that produces deterministic, fail-closed results
  • Architecture decision records and automated checks that enforce technical decisions
  • Clinical, healthcare, veterinary, or other complex domain data environments

Preferred Education and Experience

  • Formal degree requirements are not specified for this role.
  • A minimum number of years of experience is not required.
  • Candidates should demonstrate substantial practical experience in either AI development tooling or data architecture, supported by strong reasoning and the ability to deliver independently.
  • Experience with the organization’s exact technology stack is not required.
  • Machine learning or veterinary experience is not required.

Key Responsibilities

AI Development Tooling

  • Build architecture checks that compare source-tree versions across distinct dimensions and provide supporting evidence rather than a single score.
  • Develop static analysis for Python code, including checks for duplication, dependency cycles, boundary violations, exception-handling risks, and other structural issues.
  • Create reproducible, fail-closed tools that generate identical reports from identical inputs.
  • Ensure checks that cannot run never count as passing.
  • Connect architecture decision records to enforceable gates with verifiable identities, statuses, and references.
  • Generate types, constants, and validators from a single source of truth using schema-driven development.
  • Create golden fixtures that document expected behavior.
  • Develop reproducible model-graded probes and audits while making limitations and unevaluated cases explicit.

Data Architecture

  • Characterize clinical data sources by separately establishing what exists, what is populated, what can be linked, and what may legitimately be used.
  • Create versioned, per-table contracts that generate reference views, inventories, coverage reports, and data explorers.
  • Define reproducible measurements, including grain, time windows, deduplication, exclusions, unresolved cases, populations, and denominators.
  • Assess whether populations and denominators are comparable before results are compared.
  • Design safe approaches for analyzing large-scale data, including estimating query costs before accessing multi-terabyte tables.
  • Return aggregates or metadata instead of records when appropriate.
  • Evaluate existing components against requirements and reason carefully about cross-system identity.
  • Assess logical models, storage, and serving as related but separate architectural decisions.

How You’ll Work

  • Use AI agents throughout development while keeping product ownership, architecture, engineering, requirements, and implementation responsibilities distinct.
  • Define product behavior and technical decisions before implementation.
  • Document requirements, behavior, architecture decisions, detailed specifications, and implementation plans in proportion to the change.
  • Treat agents as sources of investigation and recommendations; people retain approval and authorization responsibilities.
  • Use independent reviews, source checks, and executable tests to verify that instructions are clear, claims are supported, and results meet requirements.
  • Stop affected work when decisions are missing, evidence conflicts, or instructions are unclear until the issue is resolved or an authorized temporary deviation is documented.
  • Verify completion through recorded evidence that all required checks have passed.

What Success Looks Like

  • First 30 days: Ship a small, useful check, measurement, or contract with documentation that makes the work reviewable.
  • By 90 days: Own a defined area and become a trusted resource for it.
  • By the end of the engagement: Leave behind work that is adopted, documented, and maintainable by the team.

Other Requirements

  • Work arrangement: Fully remote.
  • Time zone alignment: Eastern Time (EST).
  • Engagement: Six-month contract.
  • Employment type: W2 only.
  • Start: As soon as possible.
  • Depth in either AI development tooling or data architecture, combined with strong reasoning and curiosity about the other area, is sufficient to apply.

DeWinter Group and Maris Consulting  is an equal opportunity employer, and all qualified applicants will receive consideration for employment without regard to race, color, religion, age, sex, national origin, disability status, genetics, protected veteran status, sexual orientation, gender identity or expression, or any other characteristic protected by federal, state or local laws. We post pay scales which are based on our client pay ranges. DeWinter, Maris, and our clients have the right to modify the requirements of the role which can impact the pay ranges posted.

APPLY NOW

Share this job

Interested in this job?
Save Job
Create As Alert

Similar Jobs