{"id":4972,"date":"2025-03-11T07:52:10","date_gmt":"2025-03-11T07:52:10","guid":{"rendered":"https:\/\/www.systor.org\/2025\/?page_id=4972"},"modified":"2026-09-08T15:08:58","modified_gmt":"2026-09-08T15:08:58","slug":"isai","status":"publish","type":"page","link":"https:\/\/www.systor.org\/2026\/isai\/","title":{"rendered":"ISAI The Israeli Systems & AI Workshop 2026"},"content":{"rendered":"

[vc_row tm_bgimagefixed=”” css=”.vc_custom_1787063858914{padding-bottom: 20px !important;}”][vc_column][vc_column_text]<\/p>\n

ISAI 2026<\/h1>\n

The 3rd Israeli Systems & AI Workshop<\/h1>\n

September 7, 2026<\/h2>\n

Hosted at RedHat Raanana<\/h3>\n

<\/a>[\/vc_column_text]

Registration<\/h2>\n<\/div>
<\/span><\/div><\/header>
ISAI 2026 attendance is free of charge but requires\u00a0registration<\/a>\u00a0ahead of time and adhering to our basic\u00a0code of conduct<\/a>.<\/div><\/div>
\r\n\t\t
Register<\/a><\/div><\/div><\/div><\/div><\/section>[\/vc_column][vc_column][vc_column_text]<\/a><\/p>\n

Program<\/h2>\n

[\/vc_column_text][vc_raw_html]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[\/vc_raw_html][\/vc_column][\/vc_row][vc_row tm_bgimagefixed=”” break_in_responsive=”991″ tm_zindex=”2″ full_width=”stretch_row” tm_responsive_css=”60233685|colbreak_no|||||||||colbreak_no|||||55px||25px||colbreak_no||||||||||colbreak_no|||||||||” css=”.vc_custom_1787063745452{padding-top: 0px !important;padding-bottom: 0px !important;}”][vc_column reduce_extra_padding=”991″ tm_responsive_css=”22114509|colbreak_no|||||||||colbreak_no|||||||||colbreak_no||||||||||colbreak_no|||||||||” css=”.vc_custom_1626073647852{padding-right: 100px !important;}”][vc_raw_html]JTNDYSUyMG5hbWUlM0QlMjJJU0FJLWtleW5vdGUtYW5jaG9yJTIyJTIwc3R5bGUlM0QlMjJwb3NpdGlvbiUzQSUyMGFic29sdXRlJTNCJTIwJTIwbGVmdCUzQSUyMDBweCUzQiUyMCUyMHRvcCUzQSUyMC0xMDBweCUzQiUyMiUzRSUzQyUyRmElM0U=[\/vc_raw_html]

Keynotes<\/h2>\n[\/vc_column][\/vc_row][vc_row tm_bgimagefixed=”” css=”.vc_custom_1787063895137{padding-bottom: 20px !important;}”][vc_column width=”1\/2″]

Keynote<\/h4>\n

Towards Systems that Autonomously Evolve<\/h2>\n<\/div>
<\/span><\/div><\/header>
\n

Tamar Eilam (IBM)<\/strong><\/em><\/h3>\n

Slides<\/a><\/p><\/div><\/div><\/div><\/div><\/section><\/div> [vc_single_image image=”5235″ style=”vc_box_circle_2″][\/vc_column][vc_column width=”1\/2″][vc_tta_accordion color=”grey” c_icon=”chevron” c_position=”right” active_section=”1″ collapsible_all=”true”][vc_tta_section title=”Abstract” tab_id=”1629234536160-db282339-36e7″][vc_column_text]The emergent ability of LLMs to reason, hypothesize, experiment, and generate code is forcing us to rethink not just how we develop code, but what we design, and how we continuously evolve entire software systems.<\/p>\n

Modern systems are under increasing pressure to meet new requirements driven by a rapidly changing technological landscape such as new model architectures, hardware innovations, and shifting workloads. Yet their improvement remains fundamentally human-mediated, one decision at a time, and therefore largely reactive. At IBM Research we challenged ourselves with a question: can AI become the primary agent of system evolution, so that a deployed system observes its own behavior, hypothesizes improvements, experimentally validates them, and deploys, continuously and at machine speed?<\/p>\n

In this talk I will describe the principles we uncovered, the tools we built, and the proof points we obtained across several systems domains: inference serving optimization, accelerator compute kernel generation, and a domain-specific storage system. I will situate this within the fast-evolving academic landscape and conclude with some key open questions.[\/vc_column_text][\/vc_tta_section][vc_tta_section title=”Bio” tab_id=”1629234536247-1ed2440d-7944″][vc_column_text]Tamar Eilam received her Ph.D. in Computer Science from the Technion in 2000 and joined the IBM T.J. Watson Research Center as a Research Staff Member that same year. She was named an IBM Fellow in 2014 and her main research interest is the use of AI to manage the entire lifecycle of software systems including algorithmic discovery and optimization.[\/vc_column_text][\/vc_tta_section][\/vc_tta_accordion][\/vc_column][\/vc_row][vc_row tm_bgimagefixed=”” css=”.vc_custom_1787063922998{padding-top: 60px !important;}”][vc_column][vc_column_text]<\/a><\/p>\n

Talks<\/h2>\n

Efficient and Robust Deep Learning \u2013 Session A<\/h3>\n

Chair: Gala Yadgar <\/strong>(Technion)<\/p>\n

    \n
  • Highlight: Attention Needs Less Memory: From LLM Training to Long-Context Inference
    \n<\/b>Presented by: Malik Khalaf<\/strong> (Technion)<\/span><\/strong>
    \n    CoRR2025<\/a>
    \n    Slides<\/a><\/li>\n
  • Highlight: GASLITEing the Retrieval: Exploring Vulnerabilities in Dense Embedding-based Search<\/strong>
    \nPresented by: Matan Ben Tov<\/strong> (Tel-Aviv University)<\/span><\/strong>
    \n    CCS2025<\/a><\/li>\n
  • Highlight: Efficient Optimization of Massively Overparameterized Deep Neural Networks<\/strong>
    \nPresented by: Tom Tirer\u00a0<\/strong>(Bar-Ilan University)
    \n    TMLR2026<\/a>, NeurIPS 2025<\/a>
    \n    Slides<\/a><\/li>\n<\/ul>\n

    AI Factory Systems \u2013 Session B<\/h3>\n

    Chair: Eran Gilad <\/strong>(Regatta Data)<\/p>\n

      \n
    • What Will We Build with 100K AI Accelerators and how can You Participate
      \n<\/b>Guy Altagar<\/strong> (The National AI Program)
      \n      Slides<\/a><\/span><\/strong><\/li>\n
    • Beyond the Vector Retrieval Trap: Implementing True Episodic Agentic Memory<\/strong>
      \nGuy Korland<\/strong> (FalkorDB)
      \n      Slides<\/a><\/span><\/li>\n<\/ul>\n

      Optimizing AI Systems at Scale \u2013 Session C<\/h3>\n

      Chair: Anastasia Braginsky <\/strong>(RedHat)<\/p>\n

        \n
      • From Models to Manifests: Automating LLM Deployment Planning for llm-d on Kubernetes<\/strong><\/span>
        \n<\/strong>Amit Oren<\/strong> (Red Hat)
        \n        Slides<\/a><\/span><\/li>\n
      • The Evolution of AI: A Compute Story<\/strong><\/span>
        \nAdi Fuchs <\/strong>(Nvidia)
        \n        Slides<\/a><\/li>\n
      • AI Factory Fabric landscape and directions<\/strong><\/span>
        \n<\/strong>Shahar Belkar<\/strong> (Toga Networks)
        \n        Slides<\/a><\/span><\/li>\n
      • Highlight: LLM KV-cache: I\/O-aware KV-cache Management<\/strong>
        \nErez Zadok<\/strong> (Stony Brook University)
        \n        HotStorage2026<\/a>
        \n        Slides<\/a><\/li>\n<\/ul>\n

        <\/h3>\n

        Poster Session<\/h3>\n

        Chair: Sarel Cohen (Reichman University)<\/p>\n

        For accepted posters,\u00a0see<\/a>[\/vc_column_text][\/vc_column][\/vc_row][vc_row tm_bgimagefixed=”” css=”.vc_custom_1787064338268{padding-top: 0px !important;}”][vc_column][vc_column_text]<\/a><\/p>\n

        Venue<\/h2>\n

        Infinity Tower, 15th floor (Reception) – 8 Hapnina St. Raanana<\/p>\n

        RedHat Raanana is a convenient 5 min walk from the train station<\/strong>.<\/p>\n

        For guest parking<\/strong> – register<\/a> (after registering to ISAI) and then use the Plaza Tower guest parking entrance & lot (entrance from the HaBedolach St.)<\/p>\n