Themis AI | LinkedIn (original) (raw)
About us
Themis AI is a software development company addressing critical obstacles to the responsible deployment of AI, such as unreliable outputs and hallucinations. The solution: Capsa, a proprietary product platform that automatically adds compute-efficient uncertainty quantification to any AI model, allowing software developers across the globe to de-risk their models. Founded in 2021 by Dr. Daniela Rus, Dr. Alexander Amini, and Elaheh Ahmadi, Themis AI is an MIT Computer Science and Artificial Intelligence Laboratory spinoff grounded in more than ten years of foundational research. Driven by a team of world-class machine learning experts, Themis AI developed Capsa to equip AI professionals with the feedback they need to de-risk their models. Capsa can surgically update the architecture of any model within seconds so that the model can quantify its own uncertainty. With Capsa, AI models can flag their own mistakes and hallucinations, unlocking AI’s potential across new industries while enabling quality assurance. Capsa makes developing and running safe AI models less expensive and more efficient, saving on costly AI chips and cloud computing and decreasing time to market. As Themis AI continues to grow, it aims to make its state-of-the-art advancements in safe and reliable AI accessible to machine learning engineers and customers around the world. AI is a forecast machine to a fault. How confident should you be in the AI that serves you and your customers? Themis AI’s Capsa can tell you.
Industry
Software Development
Company size
11-50 employees
Headquarters
Cambridge, MA
Type
Privately Held
Founded
2021
Specialties
AI and ML
Locations
Employees at Themis AI
Updates
-
1,729 followers
11mo Edited
Capsa makes any AI model aware of what it does not know, all in a matter of seconds. Detect failures and improve model reliability with an industry-leading level of generality — across any model, data, or task. In this video we show how to use Capsa to (i) improve the accuracy of a state-of-the-art MNIST model, giving it the ability to identify challenging inputs during the process; (ii) make the depth estimator in an autonomous vehicle aware of high-risk scenarios it wasn't trained to handle; (iii) automatically convert a large language model into an uncertainty-aware variant that knows when it's hallucinating. Watch here: https://lnkd.in/etYWrnxB
Join now to see what you are missing
Similar pages
- VIVA AI Software Development San Francisco, California
- Liquid AI Information Services Cambridge, Massachusetts
- SymphonyAI Software Development Palo Alto, California
- Symbotic Automation Machinery Manufacturing Wilmington, MA
- Themis Software Development New York, NY
- Scale AI Software Development San Francisco, California
- Saronic Technologies Defense and Space Manufacturing Austin, TX
- Wayve Software Development London, United Kingdom
- Ontologic Data Infrastructure and Analytics Cambridge, MA
- Skydio Aviation and Aerospace Component Manufacturing San Mateo, California
Browse jobs
- Software Test Engineer jobs 16,099 open jobs
- Specialist jobs 849,142 open jobs
- Senior jobs 1,142,468 open jobs
- Python Developer jobs 52,296 open jobs
- Developer jobs 344,797 open jobs
- Policy Specialist jobs 55,689 open jobs
- Graduate Analyst jobs 60,674 open jobs
- Data Analyst jobs 550,881 open jobs
- Director jobs 1,374,979 open jobs
- Analyst jobs 760,055 open jobs
- Vice President of Quality jobs 5,494 open jobs
- Writer jobs 32,916 open jobs
- Consultant jobs 820,324 open jobs
- Associate jobs 1,094,512 open jobs
- Java Software Engineer jobs 42,983 open jobs
- Engineer jobs 608,159 open jobs
- Investment Associate jobs 43,599 open jobs
- Enterprise Sales Representative jobs 20,207 open jobs
- Director of Product Management jobs 21,042 open jobs
- Senior Product Manager jobs 63,909 open jobs