About MathMinds AI Academy is an AI-focused education technology company based in Woodbridge, VA dedicated to making personalized, high-quality STEM education accessible to students in grades 6-12. We offer in-person STEM classes and programs across mathematics, computer science, Python, and app development. Our core product is a research-driven, AI-powered learning platform that personalises each student's experience based on their performance, learning pace, and knowledge gaps. We are a small, engineering-driven team and this role sits at the center of everything we build. Our curriculum is grounded in computational thinking, real-world problem solving, and applied AI, designed to give the next generation the skills to build with technology, not just use it. You will be the primary engineer on the AI platform that students use every session. Job Overview We are looking for an AI Engineer to own the design, development, and deployment of the machine learning systems and AI infrastructure that power our personalised learning platform. This is a hands-on, full-stack AI engineering role. You will work across model development, agentic system design, backend infrastructure, and LLMOps.

You will own systems end-to-end:

from training and evaluation through production deployment, observability, and continuous improvement. This role requires the direct and sustained application of advanced knowledge in computer science, machine learning, software engineering, natural language processing, algorithms, and distributed systems.

RESPONSIBILITIES

Generative AI, LLM Systems & Agentic Engineering

• The core AI stack includes LLM-powered applications across question generation, adaptive feedback, and personalised learning recommendation systems, built on LangChain, LangGraph, and OpenAI / Anthropic APIs
• Architect and maintain multi-agent systems with tool calling, memory management, planning, and inter-agent communication; implement agentic workflows for automated content generation, assessment, and student progress analysis
• RAG pipeline ownership end-to-end: document ingestion, embedding models, vector search (pgvector, Pinecone, or equivalent), hybrid retrieval, relevance scoring, and feedback loops
• Prompt engineering, version control, A/B testing, regression detection, cost optimisation, and safety evaluation are considered basic AI principles for this role
• LLM evaluation infrastructure: task-specific metrics, offline eval harnesses, hallucination and failure mode detection, and model performance benchmarks maintained over time
• Evaluate models (GPT-4o, Claude, Gemini) for specific tasks; make principled model selection decisions based on accuracy, latency, cost, and safety
• Develop automated workflows for model evaluation, deployment, and monitoring using containerised infrastructure and orchestration tools Backend Engineering, Infrastructure & MLOps
• Architect and maintain scalable backend systems using Python, FastAPI; design clean RESTful APIs that connect AI models with the student-facing platform and instructor dashboards
• Build and manage CI/CD pipelines for model versioning, deployment, and automated rollback; implement LLMOps practices including model registries, experiment tracking (MLflow or equivalent), and deployment automation
• Implement production observability: structured logging, latency monitoring, token usage tracking, error rate dashboards, and drift/regression detection across all AI services
• Design data pipelines for ingesting, processing, and embedding educational content; manage vector stores and ensure retrieval quality at scale
• Apply responsible AI practices: implement guardrails, content safety filters, hallucination mitigation, and bias monitoring across student-facing AI features
• Implement unit, integration, and regression testing pipelines for AI services and backend APIs.
• Deploy and manage AI services using cloud infrastructure and containerized environments (AWS/GCP, Docker, Kubernetes). Applied Research & Data Analysis
• Analyze student learning data using NLP and statistical methods to identify knowledge gaps, surface learning patterns, and drive platform improvements
• Build ML models for student performance assessment, adaptive difficulty calibration, and learning outcome prediction using scikit-learn, PyTorch, or TensorFlow
• Research emerging AI techniques, agentic frameworks, and education technology approaches; prototype integrations and document findings

REQUIRED QUALIFICATIONS

• M.S. in Computer Science, Machine Learning
• directly required for this position
• Strong Python engineering skills; proficiency with PyTorch, TensorFlow, and scikit-learn
• Hands-on production experience with

LLMs:

prompt engineering, fine-tuning, RAG system design, and LLM evaluation

• Experience building agentic AI systems using LangChain, LangGraph, AutoGen, LlamaIndex, or equivalent frameworks
• Backend engineering fundamentals: FastAPI or equivalent, Postgre

SQL, REST API

design, containerization (Docker), cloud deployment

• Familiarity with MLOps / LLMOps practices: experiment tracking, model versioning, CI/CD pipelines for ML, and production monitoring
• Working knowledge of vector databases and semantic search (pgvector, Pinecone, Weaviate, or similar)
• Solid graduate-level foundations in algorithms, data structures, probability and statistics, and software systems Join us if you're eager to push the boundaries of artificial intelligence!

Bring your passion for innovation into a role where your skills will shape the future of intelligent systems in Education Technology! To apply, email us your resume at info@mathmindsai.com and fill out this form: https://forms.gle/k4b6YJ7ohRXgZymv6 We look forward to working with you!

Pay:

$20.00

• $30.

00 per hour

Benefits:

Flexible schedule

Work Location:

Hybrid remote in Woodbridge, VA 22193