PeopleofSuperintelligence

Builder

Hassan Akbari

Hassan Akbari is an AI researcher and founder whose work spans brain-speech reconstruction, image-language grounding and multimodal models. At Columbia, he coauthored an intelligible speech-reconstruction system and a CVPR paper on image-phrase grounding. He went on to coauthor VATT, PaLI, Integrated Multimodal Perception and VideoPoet; VideoPoet received an ICML 2024 Best Paper Award. As of October 2026, he leads Sciforium, an AI infrastructure company founded in 2024, as its founder and CEO after serving as an infrastructure lead at Google DeepMind.

Now at
Sciforium
Badges

Timeline

  1. 2019

    Reconstructing speech from brain signals

    At Columbia, he coauthors a deep-learning system that reconstructs intelligible speech from auditory-cortex signals. In 2019, the model improves digit intelligibility by 65% over a linear-regression baseline.

  2. 2019

    Grounding phrases in images

    In 2019, he coauthors a CVPR paper that learns a multi-level shared semantic space for visual and textual features to ground phrases in images.

  3. 2021

    VATT for video, audio and text

    He coauthors VATT, a convolution-free Transformer framework for learning multimodal representations from unlabeled video, audio and text. In 2021, the paper appears at NeurIPS.

  4. 2022

    Columbia PhD defense

    In January 2022, he defends his PhD at Columbia under Shih-Fu Chang.

  5. 2023

    PaLI language-image model

    At ICLR 2023, he coauthors PaLI, a language-image model that generates text from visual and textual inputs and performs multimodal tasks across many languages.

  6. 2023

    Integrated Multimodal Perception

    He coauthors Integrated Multimodal Perception, which combines image, video, text and audio in a single Transformer encoder. In 2023, its design combines alternating gradient descent with a mixture of experts for model and task scaling.

  7. 2024

    A VideoPoet Best Paper Award

    He coauthors VideoPoet, a decoder-only language model for zero-shot video generation from multimodal inputs. In 2024, VideoPoet receives an ICML Best Paper Award.

  8. As of October 2026

    Founder and CEO at Sciforium

    He leads Sciforium, an AI infrastructure company founded in 2024, as its Founder and CEO. He previously serves as an infra lead at Google DeepMind; at AMD's Advancing AI 2026, he appears in the speaker roster as Sciforium's founder and CEO.

Career

  1. Since 2024SciforiumFounder and CEO
  2. Google DeepMindInfrastructure lead

Education

  1. 2022Columbia UniversityPhD

From the biography.

  1. 11% of AI doctorate holders from 2010 to 2016 founded or led a company or lab by 10 years after graduationIncludes Hassan Akbari. We estimate how often AI doctorate holders founded or led a company or lab by 10 years after graduation, how waiting times compare between founders and top jobs, and which universities trained the most.
  2. Google had the most people shaping AI every year from 2014 to 2025Includes Google. Our estimates of which organizations employed the most people shaping AI each year from 1955 to 2025, tracing the shift from university labs to companies.
  3. Google has the highest combined h-index among AI organizationsIncludes Google. We estimate that Google holds the highest combined h-index at 8,016, while universities account for 73% of the deduplicated total across AI organizations.
  4. 20% of people at companies who shape AI also hold a university role, led by Google and Stanford UniversityIncludes Google. We estimate that 539 of the 6,766 people with a known current organization (8%) hold both a university role and a company role, including 20% of people at companies and 13% of people at universities.

Sources and credits

Photo: We have no published photo of Hassan Akbari yet. Send us a free-licensed photo.

  1. Towards reconstructing intelligible speech from the human auditory cortex nature.com. Checked 21 October 2026.
  2. Multi-Level Multimodal Common Semantic Space for Image-Phrase Grounding openaccess.thecvf.com. Checked 21 October 2026.
  3. CS Researchers Present at NeurIPS 2021 cs.columbia.edu. Checked 21 October 2026.
  4. Curriculum Vitae - Carl Vondrick cs.columbia.edu. Checked 21 October 2026.
  5. PaLI: A Jointly-Scaled Multilingual Language-Image Model research.google. Checked 21 October 2026.
  6. Alternating Gradient Descent and Mixture-of-Experts for Integrated Multimodal Perception research.google. Checked 21 October 2026.
  7. ICML 2024 Awards icml.cc. Checked 21 October 2026.
  8. Company sciforium.ai. Checked 21 October 2026.
  9. Build AI Systems That Scale at Advancing AI amd.com. Checked 21 October 2026.