Sarthak Jain is a member of the machine learning team at Profluent, which builds AI models for protein design. Jain is the first author of "E1: Retrieval-Augmented Protein Encoder Models", the preprint behind the openly released Profluent-E1 models in 150M, 300M and 600M parameter sizes, and a co-author of the ProGen3 scaling preprint. Profluent's team page says Jain previously worked at Amazon Research on foundation models for text and has published work on machine learning interpretability.1234
At a glance
- Current role
- Machine learning team member
- Company
- Profluent
- Sector
- Company
- Work listed
- 3
- Sources cited
- 4
- Profile reviewed
Profluent-E1
Jain is the first author, with Joel Beazer, Jeffrey A. Ruffolo, Aadyot Bhatnagar and Ali Madani, of "E1: Retrieval-Augmented Protein Encoder Models", posted to bioRxiv in November 2025. Profluent released E1 the same day as its first retrieval-augmented protein encoder family, in 150M, 300M and 600M parameter variants, available for research and commercial use.23
E1 can take homologous protein sequences as context alongside a sequence of interest, alternating between intra-sequence and block-causal attention. Profluent describes uses including fitness prediction, structure prediction and representation learning.3
Other work
Jain is the second-listed author of "Scaling unlocks broader generation and deeper functional understanding of proteins" (April 2025), the preprint accompanying Profluent's ProGen3 protein generation models. Before Profluent, Jain worked at Amazon Research on foundation models for text.41
Work and contributions
Profluent-E1 150M Open weights
First author of the E1 preprint describing the retrieval-augmented protein encoder family that includes the 150M-parameter model.23
Profluent-E1 300M Open weights
First author of the E1 preprint describing the retrieval-augmented protein encoder family that includes the 300M-parameter model.23
Profluent-E1 600M Open weights
First author of the E1 preprint describing the retrieval-augmented protein encoder family that includes the 600M-parameter model.23
Timeline
Other documented roles
Roles the sources document that are not confirmed as current.
- Researcher on foundation models for text, Amazon Research1
Where they work
- Profluent 3 people in the wiki
Related people
- Jeffrey A. RuffoloAlso at Profluent
- Ali MadaniAlso at Profluent
- Hoifung PoonAI for science · Language models
- Nick FrosstLanguage models · Interpretability
- Hendrik StrobeltInterpretability · Language models
- Jakob UszkoreitLanguage models · AI for science
- Arthur ConmyInterpretability · Language models
- Sameer SinghInterpretability · Language models
- David BauInterpretability · Language models
- Alex RivesAI for science · Language models
Suggest a correction or removal
Profiles cover public professional work only, and each statement cites a source. If something is out of date, missing or wrong, tell us and link a source. People can also ask for their profile to be corrected or removed.
References
4 sources · profile reviewed Oct 22, 2026 · data updated Oct 22, 2026