Solutions Architect, AI/ML

Fidus Systems

1d

About The Position

Fidus is seeking a Solutions Architect, AI/ML to assess technical project requirements for existing and prospective clients and determine whether opportunities are a good fit. For suitable opportunities, the role develops a solution path, project plan and estimate, and identifies risks and mitigation strategies. The Solutions Architect writes technical proposal content, explains Fidus’ capabilities and approach, and discusses, presents, defends, and refines proposed approaches and estimates with clients.

The position combines client-facing solution architecture with project work as required. It also contributes to identifying potential Fidus products and new areas for investment and research and development. The role works across Fidus’ multidisciplinary electronic product design teams and a range of technology markets, including communications, aerospace and defence, life sciences and medical devices, space and satellite, industrial automation, automotive, and semiconductors.

The position is based in the United States and involves travel, typically within North America, for up to six weeks per year. The work may require access to export-controlled information. Candidates must be a U.S. person as defined under 22 CFR 120.62 or eligible for an export license, and must complete background checks. Candidates able to undertake designs involving ITAR or classified information in the United States, or Controlled Goods Program restrictions or classified information in Canada, are sought.

Requirements

• Bachelor’s degree in Electrical Engineering, Computer Engineering, Computer Science, or equivalent.
• Ideally, at least seven years of design experience in an AI/ML engineering role, with significant experience deploying models into production rather than focusing solely on training.
• System-level design experience and interest, including identifying architectural approaches and tradeoffs.
• Expert knowledge of the modern machine-learning toolchain, including PyTorch or TensorFlow, ONNX, and quantization, pruning, and compilation processes for deploying trained models.
• Experience deploying inference on embedded and accelerated targets such as FPGA and adaptive SoC, GPU, or dedicated AI accelerators, and evaluating latency, throughput, power, and memory-bandwidth tradeoffs.
• Experience designing with CUDA and/or ROCm.
• Familiarity with AI workload hardware and infrastructure, including high-speed interconnect, memory hierarchy, and thermal and power budgets at the edge, in the data center, or both.
• Ability to assess whether a client’s AI objective is feasible given available data and to realistically scope data-pipeline, training, and validation work.
• Some knowledge of FPGAs, hardware, and embedded software.
• Ability to undertake designs involving ITAR restrictions or classified information in the United States, or Controlled Goods Program restrictions or classified information in Canada.
• Concise communication and writing skills.
• Excellent listening, understanding, and extrapolation skills.
• Professional and personable demeanor.
• Willingness to travel, typically within North America, for up to six weeks per year.
• Proven experience giving presentations to small-to-medium-sized groups.
• For access to export-controlled information, must be a U.S. person as defined under 22 CFR 120.62 or eligible for an export license; background checks are required.

Nice To Haves

• Graduate degree in machine learning or a related field.
• Exposure to MLOps practices, including model versioning, production monitoring, and retraining.

Responsibilities

• Work with prospective and existing clients to understand their needs and develop requirements.
• Architect solutions that meet client needs.
• Develop delivery plans, identify risks and mitigation strategies, and estimate the work required for client solutions.
• Write technical content for proposals.
• Explain and present Fidus’ resource capabilities and work approach.
• Discuss, promote, defend, and refine work approaches and estimates as needed.
• Evaluate solutions and divide programs into logical segments.
• Undertake project work as required.
• Help identify and develop new Fidus products.
• Help identify new areas for investment and research and development.

Benefits

• Medical plan
• Dental plan
• Vision plan
• Matching 401(k) contributions
• Annual profit-sharing bonus
• Other supplementary benefits

What This Job Offers

• Job Type: Full-time
• Career Level: Senior