Field Applications Software Engineer

Synthara; Zürich; Hybrid; CHF 120,000 - 180,000; Full time

Job summary

Synthara in Zürich invites a hands-on engineer to join our applications team, deploying ComputeRAM-enhanced workloads for AI, signal-processing and cryptography on customer systems. You'll work at the intersection of ML, embedded systems and hardware design, analysing workloads, benchmarking performance and guiding product development while supporting customers from proof-of-concept to deployment.

Excellent communication and cross-functional collaboration with hardware, software, management and

Qualifications

• Strong understanding of machine learning, deep learning and AI architectures, with practical experience using frameworks such as PyTorch or TensorFlow.
• Good understanding of SoC and computer architectures, or embedded SW
• Clear technical writing and communication skills.
• Comfortable reading technical papers, model implementations, hardware documentation and specifications and translating them into working implementations.
• Comfortable in dealing with ambiguity, working across disciplines and discussing technical requirements directly with customers, partners and internal engineering teams.
• Strong interpersonal skills and confidence engaging directly with customers, partners and cross-functional teams.
• Degree in computer science, electrical/electronic engineering, machine learning, applied mathematics, physics, engineering or a related quantitative field.

Responsibilities

• Implement and benchmark AI, ML, and signal-processing applications on ComputeRAM-enhanced systems.
• Develop proofs-of-concept for prospective customers and partners, taking workloads from initial model analysis through implementation, benchmarking and presentation of results.
• Collaborate with hardware, architecture and software teams to identify missing modelling capabilities and product features, and feed those requirements into the technical roadmap.
• Produce clear technical documentation, application notes, benchmark reports, presentations and other technical collateral.
• Maintain and expand internal understanding of ComputeRAM performance and tools.
• Develop usable simulation tools to explore new hardware design directions and generate concise reports.
• Create developer and customer documentation for benchmarks and results that are easy to understand and reproduce.

Skills

ML/AI architectures

PyTorch

TensorFlow

SoC architectures

embedded SW

technical writing

customer communication

cross-functional collaboration

Education

Bachelor's degree in CS/EE/ML

Job description

Location: Zürich, Switzerland
Seniority: 3+ years relevant industry/research experience, or equivalent advanced technical background

Join Synthara’s applications team and help demonstrate and deploy ComputeRAM®s on customer systems and AI, signal-processing and cryptography workloads.

You will work at the intersection of machine learning, signal-processing, embedded systems, computer architecture, compilers, simulators and customer applications. You will analyse and implement representative workloads, map their computational operations onto ComputeRAM, benchmark their performance and energy efficiency, and use those results to guide both customer engagements and product development. You will also work closely with Synthara’s customers, supporting them in adding ComputeRAM® to their existing products.

This is a hands-on, cross-functional role. You will work with ML and AI models, as well as signal-processing and cryptography algorithms at the application level, but also reason about hardware constraints and tradeoffs, memory movement, numerical precision, system architecture, and performance. You will collaborate closely with software, applications, platform and hardware engineers, management and sales representatives, while also supporting customer-facing proofs-of-concept, workshops, pre-sales and technical discussions.

What you’ll do

• Implement and benchmark AI, ML, and signal-processing applications on ComputeRAM-enhanced systems.
• Develop proofs-of-concept for prospective customers and partners, taking workloads from initial model analysis through implementation, benchmarking and presentation of results.
• Collaborate with hardware, architecture and software teams to identify missing modelling capabilities and product features, and feed those requirements into the technical roadmap.
• Produce clear technical documentation, application notes, benchmark reports, presentations and other technical collateral.

Outcomes (first 18 months)

• Maintain and expand company internal understanding of how ComputeRAM and software tools perform.
• Simulation tools for ComputeRAM systems are maintained and usable to explore new hardware design directions and generate easy to understand, concise reports.
• Developer and customer-quality documentation in place that makes benchmark assumptions, implementations and results easy to understand and reproduce.

Requirements

• Strong understanding of machine learning, deep learning and AI architectures, with practical experience using frameworks such as PyTorch or TensorFlow.
• Good understanding of SoC and computer architectures, or embedded SW
• Comfortable reading technical papers, model implementations, hardware documentation and specifications and translating them into working implementations.
• Clear technical writing and communication skills.
• Comfortable in dealing with ambiguity, working across disciplines and discussing technical requirements directly with customers, partners and internal engineering teams.
• Strong interpersonal skills and confidence engaging directly with customers, partners and cross-functional teams.
• Degree in computer science, electrical/electronic engineering, machine learning, applied mathematics, physics, engineering or a related quantitative field.

Nice to have (constitute a plus)

• Experience with embedded AI / TinyML or deploying neural networks onto resource-constrained devices.
• Understanding of neural-network accelerators, NPUs, GPUs, DSPs, systolic arrays or in-memory computing architectures.
• Experience with DSP workloads such as audio processing, filtering, FFTs, radar or communications algorithms.
• Experience with performance modelling, architectural simulation or hardware/software co-design.