AI/ML Engineer III

Technergetics•Utica, NY

13h•$125,000 - $175,000•Hybrid

About The Position

Technergetics is looking for an AI/ML Engineer III to design, develop, and deploy advanced AI capabilities alongside a high-performing team of full-stack developers. This role centers on building production systems around foundation models, including agentic workflows, retrieval-augmented generation, and multimodal machine learning, for demanding government and commercial customers. This position is contingent upon contract award and funding.

Requirements

At minimum, three years of professional experience in machine learning or AI systems engineering, including at least one year building with large language models or other foundation models in a production setting
Strong proficiency in Python, including asynchronous programming and modern packaging and dependency management
Working knowledge of server-side development (API definitions, REST services, streaming and asynchronous services, etc.)
Fluency with containerization and deployment frameworks such as Kubernetes or Docker, including GPU scheduling and resource management for training and inference workloads
Hands-on work with at least one major cloud platform (AWS, Azure, or Google Cloud)
Comfort with Linux platforms and command-line environments
Familiarity with Continuous Delivery/Continuous Integration (DevSecOps, GitLab Pipelines, etc.)
Proficiency with automated testing in Python (pytest), including regression suites for non-deterministic model and agent behavior
Demonstrated ability to train and deploy machine learning models with PyTorch and the Hugging Face ecosystem (transformers, datasets, accelerate)
A track record of building applications on top of large language models, including prompt engineering, structured output, and context management
Fluency with agentic frameworks and patterns such as LangGraph, LangChain, CrewAI, AutoGen, Pydantic AI, or vendor agent SDKs, including multi-step tool use, planning, memory, and state management, and with integrating models, tools, and data sources through open standards such as Model Context Protocol (MCP)
Practical command of retrieval-augmented generation, including chunking and embedding strategy, vector databases (pgvector, Milvus, Qdrant, Weaviate, FAISS), and hybrid or re-ranked retrieval
Ability to design and run LLM evaluation, including task-specific benchmarks, golden datasets, LLM-as-judge methods, and tracing and observability tooling (LangSmith, Langfuse, Arize, Weights & Biases)
Command of model adaptation techniques including fine-tuning, LoRA/PEFT, quantization, and distillation, and the judgment to know when adaptation is preferable to prompting or retrieval
Proven ability to serve models in production with inference frameworks such as vLLM, TensorRT-LLM, Triton Inference Server, Ollama, or ONNX Runtime, including latency, throughput, and cost tradeoffs, as well as deployment to edge or resource-constrained environments, including on-device inference
Grounding in AI safety and assurance practices, including guardrails, input and output filtering, prompt injection mitigation, and human-in-the-loop design
Direct work with multimodal models and cross-modal embedding across text, imagery, video, audio, or geospatial data
Demonstrated leadership on technical tasks and/or technical teams, including Agile software development and leading one or more tasks to completion
Excellent communication and teamwork skills, including the ability to explain technical tradeoffs to non-technical stakeholders and customers
Only U.S. citizens are eligible to apply, per Executive Order 12968 (Access to Classified Information).

Nice To Haves

Exposure to DoD cloud and software factory environments such as Platform One, BESPIN, AF Cloud One, DAF CLOUDworks, or AWS GovCloud, and with ATO and RMF processes at IL4/IL5
Familiarity with DoD and federal AI policy and governance, including CDAO Responsible AI guidance
Work with distributed training frameworks and schedulers (DeepSpeed, FSDP, Ray, Slurm)
Knowledge of knowledge graphs, ontologies, or Resource Description Framework (RDF), particularly as applied to graph-based retrieval (GraphRAG) and grounding
Background in developing modern full-stack web applications with frameworks such as Node, React, React Native, or Django
Basic working knowledge of Go, Java, C++, or other compiled languages
Contributions to open source AI/ML projects, or published applied AI research

Responsibilities

Leading the design, development, and deployment of multi-modal machine learning architectures, including models and algorithms, to solve complex mission and business problems
Designing and building agentic systems on top of large language models for operational deployment
Implementing retrieval-augmented generation pipelines that ground model outputs against authoritative data sources
Defining and running evaluation for model and agent behavior in production
Optimizing model inference for production and edge/DDIL (denied, degraded, intermittent, and limited bandwidth) deployment scenarios
Applying AI assurance practices and documenting model limitations to support accreditation and customer review
Integrating state-of-the-art machine learning libraries, foundation models, and agent frameworks into existing software applications
Developing and maintaining data pipelines and supporting software for collecting, preprocessing, and transforming data for machine learning tasks
Designing software solutions, algorithms, and cloud architectures needed to satisfy product features and functionality defined by the product owner and other stakeholders in a production environment
Leading, coaching, and mentoring junior data scientists, engineers, and other staff
Contributing to phases of the software development life cycle, including functional analysis, technical requirements, technical design, prototyping, coding, testing, deployment, data migration, and support
Participating in daily scrums and working with the scrum master and scrum team to organize and prioritize workload through story-pointing, supporting delivery timelines and priorities
Collaborating with cross-functional teams to understand business requirements and translate them into machine learning solutions
Performing unit testing and debugging to identify and fix software defects, and contributing to code reviews with constructive feedback to peers
Staying current on new AI/ML approaches, frameworks, and industry trends
Serving as an AI subject matter expert for small teams of researchers and engineers on advanced R&D projects funded by government and/or commercial customers, and contributing to or leading proposal writing for new opportunities within your area of expertise