Develop and maintain integrations between enterprise platforms and third-party systems, ensuring reliable, governed, and auditable data movement.
Integrate the firm's internal LLM and AI orchestration layer into client and employee portals and core business systems (e.g., Salesforce, ERP, practice management, and document management platforms), building the APIs, services, and UI components that expose AI capabilities to end users.
Design and implement retrieval-augmented generation (RAG) pipelines—document ingestion, chunking, embedding, vector search, and grounding—so AI responses draw on authoritative firm data and cite their sources.
Build agentic and tool-calling workflows through the orchestration layer, including prompt templates, tool/function definitions, multi-step agent logic, and connectors (e.g., Model Context Protocol) to enterprise systems.
Manage prompts and system instructions as versioned, testable artifacts, and establish evaluation frameworks (test datasets, automated regression testing, accuracy and hallucination scoring) to validate AI output quality before and after each release.
Implement AI guardrails and governance controls, including permission-aware data retrieval, redaction of PII and confidential client data, prompt-injection defenses, content filtering, and logging of prompts and responses for audit and compliance.
Monitor AI services in production for latency, token usage and cost, error rates, and output quality drift; tune model selection, caching, and context strategies for performance and cost efficiency.
Contribute to light application architecture decisions—evaluating build-vs-buy trade-offs, selecting appropriate patterns (microservices, event-driven, serverless), and documenting architecture—while applying security best practices (secrets management, least-privilege access, encrypted data in transit/at rest) and maintaining monitoring, alerting, logging, SOPs, and runbooks.
Required Qualifications
Bachelor's degree in Computer/Information Science, Engineering, or a related field, or an equivalent combination of education and experience.
5+ years of experience in software engineering, DevOps, or a related engineering discipline.
Proficiency in Python for automation, scripting, and integration development.
Demonstrated experience designing and building RESTful APIs and enterprise integrations.
2+ years of hands-on experience building applications that integrate large language models (LLMs) through APIs or SDKs, including prompt design, structured outputs (e.g., JSON), and tool/function calling.
Experience implementing retrieval-augmented generation (RAG), including embeddings, vector databases or search indexes (e.g., Azure AI Search, pgvector, Pinecone), and document chunking strategies.
Working knowledge of AI security and responsible-AI practices, including prompt-injection risks, data leakage prevention, access controls on retrieved data, and human-in-the-loop review.
Hands-on experience with CI/CD tooling (e.g., GitHub Actions, Azure DevOps) and version control (Git).
Familiarity with cloud platforms (Azure, AWS, or GCP) and containerization (Docker, Kubernetes).
Experience working in Agile delivery environments, with strong troubleshooting and problem-solving skills.
Strong documentation and communication skills, with the ability to translate technical concepts for varied audiences and work independently with discipline and accountability.
Preferred Qualifications (Nice to Have)
Experience with AI-assisted development tools (e.g., Claude, Copilot, or similar) for code generation, documentation, or workflow automation.
Experience with AI orchestration frameworks or platforms (e.g., LangChain/LangGraph, Semantic Kernel, Azure AI Foundry, Amazon Bedrock, or Model Context Protocol).
Experience establishing LLM evaluation and observability practices (e.g., automated evaluation suites, LLM tracing tools, token and cost monitoring), and operating internally hosted or private LLM deployments, including model serving, compute capacity planning, and model version management.
Front-end or portal development experience (e.g., React, TypeScript, or Salesforce Experience Cloud/Lightning) for embedding conversational and AI-assisted features into user-facing applications.
Familiarity with AI governance frameworks (e.g., NIST AI RMF, ISO/IEC 42001) and confidentiality requirements in accounting and tax practice (e.g., AICPA professional standards, IRS Section 7216).
Light application architecture experience (evaluating patterns, designing system boundaries, producing architecture documentation), familiarity with data integration patterns (ETL/ELT, event streaming, webhooks) and enterprise middleware or iPaaS platforms, and professional services, financial services, tax, or audit industry experience.
Experience with infrastructure-as-code (e.g., Terraform), secrets management solutions, and API security standards (OAuth 2.0, OpenID Connect, API gateways).
About Unison Systems
Unison Systems is a national IT & Engineering Services firm that delivers secure, innovative technology solutions and resources for our clients ranging from the Fortune 500 to the U.S. Navy. We bring together the strongest talent and tools in the business to solve complex problems and build solutions that make a measurable impact. At Unison, we invest in our people, foster a collaborative culture, and give our team the opportunity to learn and grow their careers. Unison is an equal opportunity employer. We are committed to creating an inclusive environment for all employees, and we do not discriminate on the basis of race, color, religion, sex, national origin, disability, or veteran status.
