DORON SADEH | CONSULTING

    What I Do Best

    Tech Savvy AI/ML hands-on expert.

    I am a hands-on AI/ML consultant and technologist with over three decades of experience — coding since 1982 and solving real-world engineering challenges since 1992. Holding an M.Sc. in Natural Language Processing and Machine Learning from the Hebrew University of Jerusalem, he has built a career spanning early-stage startups, corporate R&D, and independent consulting, with a track record that includes senior roles at Intel, a founding role at P-Cube (acquired by Cisco), and multiple CTO positions. He specializes in taking complex AI and machine learning problems from concept through architecture and all the way to production, across cloud platforms and programming languages. Beyond tech, Doron hosts a generative AI podcast, edits a literary magazine, holds patents, and in his downtime writes poetry, practices Krav Maga, and rides motorcycles.

    If you are looking for my consulting portfolio, it’s here.

    If you want to review my patents list you can find it here.

    What I Studied

    2005-2008 HEBREW UNIVERSITY, JERUSALEM, M.SC. NATURAL LANGUAGE PROCESSING & MACHINE LEARNING

    Thesis: “Deeper Magic: An Unsupervised Complex Named-Entity Recognition System“. Fully completed all mandatory and optional course duties. Thesis was fully implemented, coded, written, submitted and approved, followed up by an academic paper.

    1990-1993 HEBREW UNIVERSITY, JERUSALEM, B.Sc. COMPUTER SCIENCE & PHILOSOPHY

    Skills Acquired

    • AI/Machine Learning
    • Natural language processing
    • Algorithms
    • Parallel processing,
    • Mobile app development (Android)
    • Cloud computing (multi-platforms)
    • Networking
    • Operating systems
    • Formal languages and compiles
    • Business and product development
    • Proven management-level track record

    Experience Through the Years

    INDEPENDENT DATA SCIENCE HANDS-ON CONSULTANT, 2018-
    Working with startups and corporates creating efficient scalable AI/machine learning algorithms for big data. From problem definition, analysis, architecture, through hands-on implementation and bringing to
    production. See my portfolio for more details on customers along the years,

    fractional CTO, BetterYou, 2022-

    Part-time CTO for BetterYou, the therapy platform for the AI era, developing AI assisted psychotherapy multi-modal agents.

    Podcaster, The Generators, 2023-

    Bi-weekly podcast about generative AI bleeding edge technologies in cooperation with Shaked Zychlinski.

    EDITOR IN CHIEF A COLDER LAZARUS MAGAZINE, 2018-

    • Publishing essays about tech, culture and philosophy.
    • Essays are also re-published in Hebrew in Alaxon Magazine.

    CTO SHEILDREN TECHNOLOGIES (ANTITOXIN), 2017-2018

    • Design and architecture of a mobile (Android/iPhone) large-scale cloud backend (NodeJS/python) child protection system.
    • Developed Shiledren’s (AntiToxin) world class machine learning algorithms for online pedophile detection, verbal abuse, suicidal tendencies and many more.

    CTO AND VP R&D SHINE TECHNOLOGIES (RAINBOW), 2013-2017

    • Leading the creation of a mobile (Android) behavioral Anti-Virus (AV)application from scratch doing research, machine learning
      algorithms design and feature selection, going on to implementation.
    • Hands on coding of most of the AV app.
    • Leading the creation of a (pivoted) product, an Android network based ad blocker for both browser and app ads.
    • Designed, and coded most of the app itself. Then moving the solution to the mobile carrier level.
    • Halfway along the road, been asked to fill in for the role of VP R&D while remaining as CTO.
    • Managed a few dev teams, and QA, spanning 15-7 people during the various life cycles of the company.
    • Once the company pivoted (again) to Rainbow (network level ad filtering, and ad targeting at the carrier level) I designed and architected the full Rainbow solution from scratch, and took it from an abstract idea to a fully implemented cloud product in 6-8 months.
    • Was promoted to co-founder, by the board, under Rainbow.
    • Multiple patents pending from various pivots of the company.

    FOUNDER AND CEO XNOSIS (CTROO), 2008-2013

    • Original product was a chat bot (long before the term was coined) that analyzed natural language chats in real time, extracted topics, objects and understood the main subjects of discussion then searching the web and bringing relevant content the user could embed into the conversation (e.g. videos, images, text and commercial links to products).
    • Developed a proprietary real-time NLP-based semantic engine, from concept, through full implementation, and onto cloud deployment.
    • Engine was based on a proprietary unsupervised machine
      learning algorithm.
    • Product was canceled due to chat companies’ privacy concerns.
    • Pivoting to an in-text recommendation engine on web pages. Later
      phased to a mobile Android app.
    • Mid-way along the road, recruited a co-founder and an advisory
      board, and was partially funded by the Israeli Chief Scientist and a
      small incubator.
    • Main technologies used were Java (for the cloud based engine), browser plugins, JavaScript, Android development and multiple other open-source platforms (e.g. Stanford statistical parser), and various Amazon
      cloud service (EC2, S3, SQS, etc.).
    • Consulting (during Xnosis/Ctroo I freelanced as a hands-on consultant for the following customers)
      • MICROSOFT RESEARCH ISRAEL (JULY 2010 – JAN 2011)
        Consultant to the TeraFlux EU project, on behalf of Microsoft Research. Designing and prototyping various parts of a massive data flow architecture combining hardware, software and operating system knowhow.
      • ANY.DO LTD (AUG 2011 – DEC 2011)
        Researcher for Any.Do’s platform. Designing and prototyping
        algorithms to be part of the company’s semantic engine.
      • DESIGNART (DEC 2011 – JANUARY 2012)
        Software and hardware architecture development for 4th
        generation cellular pico base stations.
      • SUPERFISH (FEB 2012 – MAY 2013)
        Head of Search Quality, responsible for improving search results, focusing on product level, and textual analysis. From developing novel algorithms, implementing proof of concept, up until and including deploying into production. Later head of the textual algorithms group both as hands-on algorithmic/development and team leader.
      • LUCIDLOGIX (FEBRUARY 2009 – MAY 2009)
        Consulted Lucid on the Windows graphics subsystem, mainly regarding DirectX newer versions (10, and 11), with a strong emphasis on graphics/operating system interaction,
        parallelism, and emerging (GPGPU) software models.
      • SUPERFISH INC (OCTOBER 2009 – JANUARY 2010) Designed and implemented a textual semantic engine for Superfish’s internal purposes. Gone through design, and implemented the engine up to a fully functional prototype, as a part time consultant, in approx. 3 months.

    SENIOR SYSTEM ARCHITECT AND RESEARCHER, INTEL 2002 – 2008

    • Research scientist in the Corporate Technology Group reporting to an Intel fellow.
    • Lead the software design of a major Intel research Media processing project worldwide.
    • Lead the software design of an Intel-wide effort in the massive parallel processing domain (later developed into the Intel Larabee GPU product).
    • Composed a large portion of the design specs, built an international team, managed it and delivered a full-fledged software model to Intel senior management.
    • A senior system architect in the Mobile Group (IDC) co-leading the
      General Purpose GPU (GPGPU) effort.

    CHIEF LANGUAGE ARCHITECT AND TOOLS & NETWORK MANAGEMENT GROUPS’ MANAGER P-CUBE (ACQUIRED BY CISCO), 1999 – 2002

    • Second employee on-board (first developer).
    • Invented P-Cube’s high-level, parallel, network-processing programming language from scratch (SML – a deep packet inspection language in layers 3-7).
    • Chief language architect representing language architecture to industry partners.
    • Designed/implemented P-Cube’s compiler, recruiting and managing compiler team.
    • Designed/lead P-Cube’s SDK team, recruiting and managing the team.
    • Designed/lead P-Cube’s distributed network management, recruiting and managing the team.
    • Major role in designing P-Cube’s run-time embedded software/hardware.
    • Managed a total of approx. 20 people across 3 teams.

    SENIOR SOFTWARE ENG. ALGORITHMIC RESEARCH, 1998 – 1999

    • Member of the VPN joint team (ARX/Cylink).
    • Responsible for global VPN network and security management system implementation.
    • Worked with both the Israeli and U.S. based (Cylink) teams.

    PRINCIPLE SOFTWARE ENG. DIGITAL EQUIPMENT CORP. 1992 – 1998

    • Joined as a QA student during BSC (FDDI project).
    • Came on board full time into the DEC 100MB PCI drivers team.
    • Took over the Novel driver certification effort, both in Israel and US.
    • Novel kernel driver’s developer responsible for both development and international customer service.
    • Initiated, implemented & lead the Novel, Unix, and NDIS drivers unified driver architecture in C++.

    When not Working

    • Write and translate Poetry/Short Stories
    • Compose essays for A Colder Lazarus & Alaxon Magazines
    • Practice carpentry
    • Krav Maga
    • Ride motorcycling, and 4×4
    • Raise (my) young kids