Please note this posting is to advertise potential job opportunities. This exact role may not be open today, but could open in the near future. When you apply, a Cisco representative may contact you directly if a relevant position opens.

Who We Are

Cisco Silicon One is the center of Cisco’s ASIC design and is driving the development of next-generation network devices for the 5G and AI era. Our outstanding team works in a startup atmosphere within a stable and leading corporate environment, and our design center hosts all silicon HW and SW development fields under one roof. Our devices are crafted for universal adaptability, delivering high speed without sacrificing programmability, buffering, power efficiency, scale, or feature flexibility.

What You'll Do

You will conduct training and inference benchmarks for machine learning jobs to ensure efficient performance. Develop a repeatable test bench for experimenting with the latest model serving infrastructures (e.g., TGI, VLLM, DeepSpeed) for both training and inference.

• Apply well-known open-source training models for fine-tuning and implement techniques like RAG (Retrieval-Augmented Generation).
• Generate training datasets crucial for model training, ensuring data quality and relevance.
• Design and implement a multi-tenant chatbot using company documents, exposing them as externally accessible chat agents.
  
  

In this role, you will be putting your expertise in distributed GPU training and inference to good use, particularly with infrastructure systems like SLURM and Kubernetes, among others. You will also regularly demonstrate your familiarity with ML accelerator frameworks such as Accelerate and DeepSpeed. A key aspect of your responsibilities will be applying your knowledge of data parallelism and tensor parallelism, along with successfully explaining these concepts.

This position offers a hybrid work arrangement. You are welcome to operate from our offices located in Caesarea, Tel Aviv, or Modi'in.

Who You Are:

• Bachelor's or higher degree in Computer Science, Machine Learning, or a related field.
• Experience (2-5 years) in running training and inference benchmarks for machine learning.
• Strong proficiency in implementing and experimenting with model serving infrastructures.
• Hands-on experience with open-source training models and fine-tuning techniques.
• Ability to create and curate training datasets for diverse machine learning models.
• Experience in distributed GPU training and inference using SLURM, Kubernetes, or similar infrastructure.
• Familiarity with ML accelerator frameworks like Accelerate, DeepSpeed, and others.
• Knowledge of data parallelism and tensor parallelism.
• Proficient in relevant programming languages (e.g., Python, TensorFlow, PyTorch).
• Excellent problem-solving skills and the ability to work in a collaborative team environment.
  
  

Preferred Skills:

• Familiarity with the latest advancements in machine learning frameworks.
• Experience with natural language processing (NLP) and chatbot development.
• Knowledge of cloud platforms and services for scalable model deployment.
• Strong communication skills to collaborate with multi-functional teams.
  
  

If you meet these qualifications and are enthusiastic about pushing the boundaries of machine learning infrastructure, we invite you to apply and contribute to our innovative projects.

Learn more about us right here:

[1] https://www.cisco.com/c/en/us/solutions/service-provider/innovation/silicon-one.html

[2] https://fortune.com/2019/12/11/cisco-ceo-chuck-robbins-chips/?_lrsc=ce750ac5-11c8-4995-bfd3-936de1c06009&dtid=osolin001080

[3] https://www.calcalist.co.il/internet/articles/0,7340,L-3775619,00.html

[4] https://www.youtube.com/watch?v=YRJuFrZC-GE&list=PLFT-9JpKjRTA0aqRbr5sR2iiaqVtK6Wf0&index=1

[5] https://blogs.cisco.com/sp/five-principles-at-the-heart-of-cisco-silicon-one?_lrsc=6c2b8b61-b513-4457-bdc9-12b9f9ecef45&dtid=osolin001080

Why cisco

#WeAreCisco, where each person is unique, but we bring our talents to work as a team and make a difference. Here's how we do it.

We embrace digital, and help our customers implement change in their digital businesses. Some may think we’re “old” (30 years strong!) and only about hardware, but we’re also a software company. And a security company. A blockchain company. An AI/Machine Learning company. We even invented an intuitive network that adapts, predicts, learns and protects. No other company can do what we do – you can’t put us in a box!

But “Digital Transformation” is an empty buzz phrase without a culture that allows for innovation, creativity, and yes, even failure (if you learn from it.)

Day to day, we focus on the give and take. We give our best, we give our egos a break and we give of ourselves (because giving back is built into our DNA.) We take accountability, we take bold steps, and we take difference to heart. Because without diversity of thought and a commitment to equality for all, there is no moving forward.

So, you have colorful hair? Don’t care. Tattoos? Show off your ink. Like polka dots? That’s cool. Pop culture geek? Many of us are. Passion for technology and world changing? Be you, with us!

We will ensure that individuals with disabilities are provided reasonable accommodation to participate in the job application or interview process, to perform essential job functions, and to receive other benefits and privileges of employment. Please contact us to request accommodation.

#WeAreCisco , :)