AI Software Engineer

📋 Description

• • Own the end-to-end creation of mission-critical Generative AI solutions that directly improve how federal agencies serve millions of Americans. From the first white-board sketch to production traffic on classified networks, you will be the technical anchor who turns cutting-edge research into reliable, secure, and scalable systems.
• • Architect and implement full-stack LLM applications using RAG pipelines, vector databases, and AI Agents that reason over terabytes of structured and unstructured government data. You will decide when to fine-tune, when to prompt-engineer, and how to chain models for optimal accuracy, latency, and cost.
• • Lead rapid, stakeholder-driven prototyping inside two-week sprints. Translate vague policy directives into concrete user stories, build clickable demos, gather feedback from agency program managers, and pivot before a single line of code becomes technical debt.
• • Engineer bullet-proof MLOps / LLMOps workflows that automate testing, bias detection, performance monitoring, and rollback. Your pipelines will run in FedRAMP-authorized Azure and AWS environments, satisfying NIST RMF controls without slowing delivery.
• • Champion Responsible AI across every commit. Embed fairness checks, privacy-preserving techniques, and explainability layers so that civil servants and citizens alike can trust the outputs of your models.
• • Harden cloud-native microservices for IL5+ security enclaves. Use Infrastructure-as-Code (Terraform/Bicep) to spin up GPU clusters, vector stores, and serverless endpoints that auto-scale with classified workloads yet cost less than a single on-prem rack.
• • Publish original research and reference architectures through Anika’s VITAL Innovation Lab. Present findings at MITRE, DEF CON, and internal Communities of Practice, influencing the next wave of federal AI policy and procurement.
• • Mentor junior engineers and data scientists. Run lunch-and-learns on LangGraph patterns, GPU kernel optimization, or the latest NIST guidelines, multiplying your impact across the company and client agencies.
• • Collaborate with cross-functional squads of policy analysts, UX designers, and cybersecurity engineers. Your code will sit beside theirs in a unified GitHub mono-repo, reviewed daily and deployed hourly via GitHub Actions.
• • Continuously scan the horizon for emerging open-source models, quantization techniques, and vector-search innovations. Run spike experiments on Fridays and, if the metrics beat the baseline, promote to production on Monday.
• • Contribute to capture efforts and proposal writing. Translate your technical wins into clear, persuasive prose that wins follow-on contracts and keeps Anika at the forefront of federal AI delivery.
• • Travel occasionally (≤10 %) to agency sites in DC, Baltimore, or contractor facilities for classified integration work. All travel is planned weeks in advance and supported by company logistics.