Alexander Amini

Alexander Amini is an American computer scientist, co-founder and chief scientific officer of Liquid AI, the 2023 MIT CSAIL spinout building Liquid Foundation Models on the liquid-neural-network architecture. He is also a research scientist at MIT CSAIL in Daniela Rus's Distributed Robotics Laboratory and the lead organizer and lecturer of MIT 6.S191: Introduction to Deep Learning, MIT's flagship deep-learning course whose lectures have been viewed more than 10 million times globally. He is a co-author on the foundational Liquid Time-Constant Networks paper that anchors Liquid AI's research thesis.

At a glance

• Education: BS in electrical engineering and computer science with a minor in mathematics, Massachusetts Institute of Technology (MIT), 2017; MS in computer science, MIT, 2018; PhD in computer science, MIT, May 2022. Doctoral dissertation: "End-to-end Learning for Robust Decision Making," advised by Daniela Rus at MIT CSAIL's Distributed Robotics Laboratory.
• Current roles: Co-founder and chief scientific officer at Liquid AI (since March 2023); research scientist at MIT CSAIL (since June 2022); co-founder and advisor at Themis AI (since 2021); lead organizer and lecturer for MIT 6.S191: Introduction to Deep Learning (since January 2018).
• Key contributions: Liquid time-constant networks; closed-form continuous-time neural networks; the VISTA data-driven autonomous-vehicle simulator; deep evidential regression; MIT 6.S191 course curriculum.
• Awards: Hyperion Research Innovation Excellence Award (2022); JP Morgan Chase Graduate Research Fellowship (2021); NSF Graduate Research Fellowship (2017); MIT Outstanding Mentor for Undergraduate Researchers (2020); IEEE ICRA Best Overall Paper Finalist (2019); Grand Prize Winner, European Union Contest for Young Scientists (2011); Grand Prize Winner, BT Young Scientist and Technologist Competition (2011).
• X: @xanamini
• LinkedIn: Alexander Amini
• Personal site: mit.edu/~amini
• Google Scholar: Alexander Amini

Origins

Amini was born on April 29, 1995, in Fort Worth, Texas, and grew up in Yorktown Heights, New York. He began playing tennis at age five, taught by his father. After ninth grade his family moved to Dublin, Ireland, where Amini enrolled at Castleknock College in August 2010. The relocation produced his first widely-publicized research project, "Tennis Sensor Data Analysis: An Automated System for Macro-motion Refinement," which combined inertial-sensor data with classification algorithms to identify tennis stroke types and provide real-time motion-correction feedback.

In January 2011, at age 15, Amini won the BT Young Scientist and Technologist of the Year award representing Ireland, and in September 2011 the project won the top prize at the European Union Contest for Young Scientists. The tennis-sensor research produced two US patents and informed the early commercial venture TennisTek.

In August 2013 Amini returned to the United States to enroll at MIT, where he completed all three degrees in computer science (BS 2017, MS 2018, PhD May 2022). He held a research-fellow appointment at MIT's Senseable City Laboratory and completed summer internships at IBM Research (2016) and NVIDIA's end-to-end driving team (2017), where his uncertainty-estimation work was deployed on full-scale self-driving vehicles.

Career

Amini joined Daniela Rus's Distributed Robotics Laboratory at MIT CSAIL in August 2017 as a graduate researcher. The doctoral research focused on end-to-end learning for autonomous decision-making, uncertainty estimation for deep neural networks, and data-driven simulation, producing more than thirty peer-reviewed publications at NeurIPS, ICML, ICLR, ICRA, IROS, AAAI, RA-L, Nature Machine Intelligence, and Science Robotics.

In January 2018 Amini took over as lead organizer and lecturer of MIT 6.S191: Introduction to Deep Learning alongside Ava Soleimany (now Ava Amini, a principal researcher at Microsoft Research). The course was rebuilt by Amini and Soleimany, who met as undergraduates on the MIT tennis courts and subsequently married. Through eight consecutive January editions from 2018 to 2025 and the active 2026 edition, the course has reached more than 80,000 globally registered students across 55 countries, with YouTube lectures accumulating over 10 million views.

The doctoral research line produced the 2019 Variational End-to-End Navigation and Localization paper at ICRA (Best Overall Paper Finalist, top 0.1%); the 2020 Deep Evidential Regression paper at NeurIPS; and the VISTA and VISTA 2.0 data-driven simulators (RA-L 2020, ICRA 2022) for training autonomous-vehicle policies that transfer to real-world driving. The collaboration with Ramin Hasani and Mathias Lechner, the TU Wien-trained researchers who joined Rus's group at MIT CSAIL, produced the foundational liquid-network research portfolio. Amini is a co-author on the June 2020 Liquid Time-Constant Networks paper accepted at AAAI-21 with an oral spotlight, the October 2020 Nature Machine Intelligence paper "Neural circuit policies enabling auditable autonomy," the November 2022 paper "Closed-form continuous-time neural networks," and the April 2023 Science Robotics cover story "Robust flight navigation out of distribution with liquid neural networks."

In 2021 Amini co-founded Themis AI with Rus and Soleimany, an MIT CSAIL spinout commercializing uncertainty-aware AI methods. Following the May 2022 defense he was appointed a research scientist at MIT CSAIL in June 2022. The dissertation was recognized by the Hyperion Research Innovation Excellence Award.

Liquid AI was incorporated on March 30, 2023, with Hasani as chief executive, Lechner as chief technology officer, Amini as chief scientific officer, and Rus as senior co-founder and advisor. The company emerged from stealth in December 2023 and raised a $250 million Series A in December 2024 at a $2.35 billion valuation, led by AMD, with cumulative funding through April 2026 of approximately $297 million. In April 2026 Liquid AI announced a strategic partnership with Mercedes-Benz to embed LFMs into the MBUX infotainment system, with first North American rollout scheduled for the second half of 2026. Throughout the Liquid AI period Amini has retained the MIT CSAIL research-scientist appointment, the 6.S191 course leadership, and the Themis AI advisory role, and is the named first author on the November 2025 LFM2 technical report.

Affiliations

• MIT: Bachelor's, master's, and doctoral student, EECS, 2013 to 2022
• MIT CSAIL: Doctoral researcher (Distributed Robotics Laboratory, advisor Daniela Rus), 2017 to 2022; research scientist, 2022 to present
• Themis AI: Co-founder and advisor, 2021 to present
• Liquid AI: Co-founder and chief scientific officer, March 2023 to present

Notable contributions

• Liquid Time-Constant Networks. "Liquid Time-Constant Networks" (Hasani, Lechner, Amini, Rus, Grosu), accepted at AAAI-21 with an oral spotlight, introduced the LTC neural-network family that became the architectural basis for the broader liquid-network research line and Liquid AI's commercial Liquid Foundation Models.
• Closed-form continuous-time neural networks. The November 2022 Nature Machine Intelligence paper "Closed-form continuous-time neural networks" demonstrated a closed-form approximation of the LTC integral, producing models that train and run between one and five orders of magnitude faster than differential-equation-based architectures.
• VISTA data-driven simulator. VISTA, introduced in 2020 RA-L and expanded in "VISTA 2.0" (ICRA 2022, Amini and Tsun-Hsuan Wang as co-leads), supports RGB cameras, 3D LiDAR, and event-based cameras for autonomous-driving policy training with substantially less collected data than alternative approaches.
• Deep evidential regression. "Deep Evidential Regression" (Amini, Schwarting, Soleimany, Rus, NeurIPS 2020) introduced a method for deep neural networks to express both aleatoric and epistemic uncertainty without ensembles.
• Robust flight navigation with liquid neural networks. "Robust flight navigation out of distribution with liquid neural networks" (Science Robotics, April 2023, cover spotlight) demonstrated that liquid neural networks could pilot a quadrotor through unfamiliar visual environments and out-of-distribution scenes.
• MIT 6.S191. Lead organizer and lecturer of MIT's flagship deep-learning course since January 2018, alongside Ava Amini. Open-sourced YouTube lectures have accumulated over 10 million views from more than 80,000 students across 55 countries.
• TEDxMIT talk. "Intelligence that we can trust" (TEDxMIT, December 2022), Amini's public account of uncertainty-aware deep learning.

Investments and boards

The entries below are limited to AI, semiconductors, datacenters, software, and energy.

• Liquid AI (AI): Co-founder and chief scientific officer, March 2023 to present. MIT CSAIL spinout developing Liquid Foundation Models. $250 million Series A in December 2024 at a $2.35 billion valuation led by AMD. Cumulative funding approximately $297 million through April 2026.
• Themis AI (AI): Co-founder and advisor, 2021 to present. MIT CSAIL spinout commercializing uncertainty-aware AI deployment methods, co-founded with Daniela Rus and Ava Amini.

No other public investor activity on record in AI, semiconductors, datacenters, software, or energy as of May 2026.

Network

Amini's longest-running and most central professional relationship is with Daniela Rus, his MIT doctoral advisor, the senior co-founder of Liquid AI, and a co-founder of Themis AI. The Rus advising relationship dates from August 2017 and bridges every major academic and commercial milestone of his career. Ramin Hasani and Mathias Lechner, the TU Wien-trained researchers who joined Rus's group at MIT CSAIL during Amini's doctoral period, are his Liquid AI co-founders and longest-running research collaborators outside the Rus relationship. Ava Amini (formerly Ava Soleimany), a principal researcher at Microsoft Research and 6.S191 co-instructor since 2018, is his wife and a co-author on multiple papers. Sertac Karaman of MIT AeroAstro and Guy Rosman of Toyota Research Institute are recurring co-authors on the autonomous-vehicle research line. AMD's senior leadership, the lead Series A investor, is the principal commercial-investor relationship.

Position in the field

Amini occupies a structurally distinctive position among insurgent AI lab co-founders through the combination of a ten-year continuous research career at a single institution, the Daniela Rus advising and co-founding relationship, the public-facing role as lead instructor of one of the most-watched university deep-learning courses online, and the commercial trajectory of Liquid AI from 2023 spinout to a $2.35 billion Series A valuation in eighteen months.

Industry coverage has frequently characterized Amini as the autonomous-systems and product co-founder of Liquid AI alongside Hasani as CEO and Lechner as CTO. Among MIT CSAIL spinout co-founders, the Hasani-Lechner-Amini founding-team split is the operating template, with Rus as the senior academic advisor. The 6.S191 course has produced an additional public-facing role that few other insurgent-AI co-founders match in scale and continuity.

The architectural thesis underlying Liquid AI remains contested on whether liquid-network architectures scale to frontier-tier capability or remain confined to smaller-parameter and edge-deployment segments. The Liquid AI commercial trajectory through 2026 and 2027 and the LFM successor model releases will provide the principal evidence on the architectural-scaling question, with the model-research responsibility falling primarily to the chief-scientific-officer role.

Outlook

Open questions over the next 6 to 18 months:

• Liquid AI model scaling. Whether the LFM2 line and subsequent successor models scale the architectural-efficiency advantage toward larger parameter classes that compete more directly with transformer-family flagship models.
• 6.S191 continuity. Whether Amini sustains the lead-instructor role alongside the Liquid AI executive responsibilities, or whether course leadership transitions to other lecturers as the company scales.
• Autonomous-systems integration in LFM products. Whether the autonomous-vehicle and robust-decision-making research line that anchored the dissertation produces commercial features in the Liquid AI product line.
• Mercedes-Benz commercial validation. Whether the second-half-2026 MBUX rollout produces the engineering validation that converts the architectural thesis into a durable edge-AI business.
• Public-research cadence. Whether Amini continues the conference-publication cadence of the doctoral and Liquid AI period, or whether CSO and 6.S191 responsibilities reduce the bandwidth.

Sources

• Alexander Amini personal site. Personal page with research summaries, publication list, teaching, and entrepreneurship overview.
• Alexander Amini CV. Complete CV with education, experience, awards, and publications.
• Alexander Amini | Liquid AI. Liquid AI team page with current title and biographical summary.
• Alexander Amini | MIT CSAIL. MIT CSAIL profile page.
• End-to-end Learning for Robust Decision Making. May 2022 MIT doctoral dissertation, advised by Daniela Rus.
• MIT 6.S191: Introduction to Deep Learning. Course site with curriculum and lecture archive.
• Liquid Time-Constant Networks. The foundational LTC paper, accepted to AAAI-21.
• Closed-form continuous-time neural networks. November 2022 Nature Machine Intelligence paper.
• VISTA 2.0. 2022 ICRA paper introducing the multimodal autonomous-vehicle simulator.
• Deep Evidential Regression. 2020 NeurIPS paper on evidential deep learning for uncertainty quantification.
• Robust flight navigation out of distribution with liquid neural networks. April 2023 Science Robotics cover-spotlight paper.
• Intelligence that we can trust | TEDxMIT. December 2022 TEDxMIT talk on uncertainty-aware deep learning.
• Bringing deep learning to life. February 2020 MIT News profile of Amini and Soleimany on 6.S191 instruction.
• Dublin student wins EU Young Scientist award. September 2011 RTE coverage of the European Union Contest for Young Scientists Grand Prize.
• Liquid AI: A new MIT spinoff wants to build an entirely new type of AI. December 2023 TechCrunch coverage of Liquid AI's emergence from stealth.
• Liquid AI: We raised $250M to scale capable and efficient general-purpose AI. December 2024 Series A announcement.
• Photo: Liquid AI team page, Liquid AI press portrait.