Microsoft Technology/Software

Principal Applied Scientist - CTO Office

Herzliya, Israel In person Full-time

Overview

The Microsoft Security office of the CTO is expanding to drive the development of a new type of GenAI foundation architecture that can effectively address the unique challenges of cybersecurity: cybersecurity can benefit from holistic processing of the huge “haystacks” of security data, in order to better find and correlate the few “needles” of malicious behaviors; however, current LLM technology cannot handle such vast amounts of data concurrently. This new joint initiative aims to develop a new neural-network architecture that can overcome this limitation. This is a unique opportunity to engage in frontier research that is also product-focused.

You’ll join a small, multidisciplinary and technically deep team in the Microsoft Security Office of the CTO, built to take on important problems whose answers are not yet known. We bring together applied AI, research, architecture, engineering and product thinking in empowered teams with clear objectives, short learning cycles and substantial autonomy. Our job is not simply to deliver against an established roadmap: we explore emerging opportunities, build and experiment, generate evidence, validate new technologies with customers, and establish the technical direction that enables product organisations to scale what works.

It’s an environment for scientists who want the freedom to pursue genuinely hard technical questions while staying relentlessly connected to real customer and product impact.

As a Principal/Senior Applied Scientist, you will own end‑to‑end model development for security scenarios, including developing new model architectures, continual pre‑training, task‑focused fine‑tuning, reinforcement learning, and objective, benchmark‑driven evaluation to ensure models are optimized, validated, and production ready. You will also drive training efficiency and reliability on distributed GPU systems, deepen model abilities (reasoning, tool‑use, prediction horizon, etc), and embed Responsible AI, privacy, and compliance.

The role is hands‑on and impact‑focused, partnering closely with engineering and product to translate innovations into shipped, measurable outcomes, defining quality gates and readiness criteria, and mentoring scientists and engineers to scale results across globally distributed teams.

You will combine strong coding, experimentation, and debugging skills with a systems mindset to accelerate iteration cycles, improve throughput and cost‑effectiveness, and help shape the next generation of secure, trustworthy AI for our customers, from model training through applied, continuously learning security systems in production.

Responsibilities

You’ll work as part of an Applied Science team on high-impact, technically ambitious AI projects that directly shape the future of AI in Cyber security, with ownership for taking advanced research through to production impact.

Technical Leadership & Ownership: set technical direction for major security domain initiatives; lead security model programs spanning pre‑training, task tuning, reinforcement learning, and evaluation; translate cutting‑edge research into production‑ready capabilities.

Advanced Model Design – Building and customizing deep learning model architectures (e.g., modifying transformer blocks, attention/memory modules, etc.) at the SLM/LLM scale; making principled architectural tradeoffs to improve reliability, robustness, and security‑specific behavior.

Advanced Model Training – Apply deep expertise in pre-training, post-training, and reinforcement learning (RL) for both language and other modalities, including time-series.

Design & Evaluate Datasets – Build high-quality datasets and benchmarks; define objective evaluation frameworks and quality gates; run ablation studies to measure impact and optimize data and training effectiveness to support confident product decisions.

Develop Data Infrastructure – Create and maintain scalable pipelines for ingestion, preprocessing, filtering, and annotation of large, complex datasets, with attention to privacy, governance, and long‑term reuse across security scenarios and continual learning loops that improve deployed agents over time.

Research & Innovation – Collaborate with cross-functional teams to push research and product boundaries, delivering models that make a real-world impact.

M.Sc. / Ph.D. in Computer Science, Information Systems, Electrical or Computer Engineering or Data Science (Ph.D. strongly preferred). Candidates with M.Sc. / Ph.D. in related fields with proven industry experience or a strong publication record in the areas of LLM, Information Retrieval, Machine Learning, Natural Language Processing, Time Series Forecasting and Deep Learning are considered as well.

Proven hands-on experience of at least 5 years (including post-grad work) in building and deploying Machine Learning products. Key areas of expertise include Natural Language Processing and Large Language Models, along with an understanding of concepts such as Privacy and Responsible AI. Candidates are expected to demonstrate a strong history of successfully translating applied research into production-ready solutions, along with a proven track record of delivering projects within large-scale production environments.

Proven expertise in the LLM and/or time-series forecasting domain, demonstrating comprehensive knowledge of relevant concepts in the domain. Ideal applicants should be proficient in areas such as LLM’s pre and post training, including CPT, SFT and RL, LLM benchmarking, agentic flows, and model alignment.

Hands-on experience in building neural model architectures at the 100M+ scale and the proficiency to adapt them at all abstraction levels down the individual block (e.g. changing the innerworkings of an attention block, introducing new blocks, or changing the routings)

Demonstrated proficiency in problem-solving and data analysis, with substantial expertise in evaluating the performance of large language models (LLMs) and/or time-series forecasting models, developing benchmarks tailored to practical scenarios.

Preferred Qualifications

PhD degree in Computer Science, Information Systems, or Data Science.

Evidence of research contributions through publications or records of top-tier journal and conference publications or submitted/accepted papers in top venues (KDD, ICML, AAAI, ACL, ICLR, etc.)

Proven track record in pre/post-training of large transformer models for language and/or time series tasks.

Customer obsession and passion about making real world product impact through production deployed systems.

Excellent verbal and written communication skills, with the ability to simplify and explain complex ideas.

Effective collaboration skills while working within a globally distributed organization.

This position will be open for a minimum of 5 days, with applications accepted on an ongoing basis until the position is filled.

Microsoft is an equal opportunity employer. All qualified applicants will receive consideration for employment without regard to age, ancestry, citizenship, color, family or medical care leave, gender identity or expression, genetic information, immigration status, marital status, medical condition, national origin, physical or mental disability, political affiliation, protected veteran or military status, race, ethnicity, religion, sex (including pregnancy), sexual orientation, or any other characteristic protected by applicable local laws, regulations and ordinances.

If you need assistance with religious accommodations and/or a reasonable accommodation due to a disability during the application process, read more about requesting accommodations.

Insights from previous hires

Top skills: Science

Common previous titles: Applied Researcher, Applied Scintist, Applied Sciense, Cloud Solution Architect, Senior Technical Program Manager, Senior Data Analyst, Senior Solution Architect, Senior Researcher, Senior Analog Design Engineer, Security Analyst

Free tool / Resume tailor

Tailor your resume for this job

Paste your resume and get 3 to 8 of your own lines rewritten in the words this posting uses, with the reason for each change. The job is already filled in, and you don't need an account.

Job description provided by Microsoft.
View the original posting