Generative AI Software Engineer

Koniag Government Services, LLC•Washington, DC

1d•$149,000 - $175,000

About The Position

Tuknik Government Services, LLC, a Koniag Government Services company, is seeking a Generative AI Software Engineer to support a U.S. Army customer on a large-scale contracting systems modernization and automation program. The engineer will build secure generative AI services and workflow integrations under approved architecture and Responsible AI controls. The work includes implementing retrieval, prompt, orchestration, evaluation, and monitoring components that assist users without replacing required Government decisions. This is a contingent recruiting profile for a future new business opportunity; the opening, employing subsidiary, compensation, work location, and staffing are subject to final contract authorization. The role requires U.S. citizenship and an active security clearance or eligibility required for assigned work by the start of performance, with ongoing eligibility thereafter. Work may occur remotely when authorized, at Government facilities, or in controlled office and data-center environments. Work involving classified information or SIPRNet must take place at a Government-approved secure location using authorized systems. The role generally supports Government business hours, with occasional release, incident-response, maintenance-window, or approved travel requirements. Reasonable accommodations may be made to enable qualified individuals with disabilities to perform the essential functions.

Requirements

Bachelor's degree in computer science, engineering, data science, or a related field; additional relevant experience may substitute for education.
Four or more years of software engineering experience, including practical development with large language models or generative AI services.
Strong Python or comparable backend development skills.
Experience with REST APIs, authentication, structured data, testing, and observability.
Experience implementing retrieval-augmented generation, prompt orchestration, automated evaluations, and secure AI application patterns.
U.S. citizenship and the active security clearance or eligibility required for assigned work by the start of performance, with ongoing eligibility thereafter.
Ability to communicate clearly with Government, technical, and business stakeholders and maintain accurate work records in approved systems.
Ability to protect Controlled Unclassified Information and classified information when assigned and comply with required access, training, and handling procedures.
Python and AI service development.
Retrieval-augmented generation and prompt engineering.
Automated model evaluation and guardrails.
Secure API integration and observability.
DevSecOps delivery and technical documentation.

Nice To Haves

Experience with AWS Bedrock, commercial or Government generative AI platforms, vector search, and agent frameworks.
Experience integrating AI into Appian or another workflow/case-management platform.
Knowledge of DoD authorization, CUI handling, Responsible AI, and software supply-chain controls.

Responsibilities

Develop generative AI services for approved uses such as document drafting, policy analysis, knowledge retrieval, summarization, and diagnostic assistance.
Implement secure integrations between Appian and authorized model endpoints using structured requests and responses.
Build retrieval-augmented generation workflows using approved authoritative sources, metadata, access controls, and citation requirements.
Develop prompt templates, orchestration logic, validation rules, guardrails, and exception-handling paths.
Create automated evaluation suites for accuracy, groundedness, citation quality, safety, latency, cost, and expected behavior.
Implement role-based access, data minimization, redaction, audit logging, model and prompt versioning, and human-approval controls.
Package and deploy AI components through approved DevSecOps pipelines and produce SBOM, test, security, and release evidence.
Monitor production quality, usage, failures, token or consumption patterns, and user feedback, and tune solutions within approved limits.
Investigate AI-related defects and coordinate corrections with Appian, data, security, API, and product teams.
Document solution architecture, data flows, prompts, evaluations, operational procedures, and known limitations.
Protect Controlled Unclassified Information and classified information when assigned, and comply with required access, training, and handling procedures.
Communicate clearly with Government, technical, and business stakeholders and maintain accurate work records in approved systems.