Dreamfold
Founded 2022 · 6 employees on LinkedIn · 3 known investors
Dreamfold engineers proteins as therapeutic medicines, using generative AI models that reason over protein sequence and structure to enable potent, localized immune modulation with minimal off-target effects.
Also known as DreamFold · DreamFold.ai · Wise Algorithms Inc.
Investors · 3
.jpeg){kind=logo}
Funding
SEC filings, press & company announcements- Undisclosed amountSeedJun 2023
IQ Capital (lead)
Source ↗
Source: company announcements and press reports — follow each round's link for the claim.
Company profile
researched Aug 2026Dreamfold is a protein design company that develops generative machine learning models to navigate protein sequence and structure space in search of new medicines. Its stated focus is engineering proteins as programmed medicines that enable potent, localized immune modulation with minimal off-target effects, aimed at cancer, autoimmune and other immune-related diseases where existing immune therapies leave patients without effective treatment. The company combines machine learning with molecular simulation and synthetic biology, and runs continuous wet-lab testing whose sparse experimental ("black-box") readouts feed back into its learned protein structure representations.
The company originated at Mila (Quebec AI Institute) and was founded by a team of Mila students and researchers, with a team spanning computational scientists, chemists and biologists. Under an earlier positioning it was described as an AI-based generative search engine for antibodies, developing a drug discovery pipeline based on high-throughput biological experimentation, active learning and GFlowNets to design antibodies targeting membrane proteins. Its legal name is recorded as Wise Algorithms Inc.
Founding story
Dreamfold's roots trace to the COVID-19 period, when Maksym Korablyov, co-founder and CEO, initiated a project at Mila to explore drug design solutions for the pandemic. That project evolved into a broader vision of applying generative modeling to protein design, and became Dreamfold, a startup formed out of Mila's research community. Jarrid Rector-Brooks is co-founder and CSO, and the team is composed of current and former Mila members. Investor materials describe the company as founded by researchers from Yoshua Bengio's lab alongside Michael Bronstein, a professor at the University of Oxford.
Traction
The team has published several papers, including spotlight presentations at ICLR and ICML, and developed FoldFlow-2. Initial protein synthesis results from its models were described as promising and a basis for further biological validation.
Latest developments
A March 2025 report described Dreamfold raising $6 million to expand computational infrastructure and its team of scientists and engineers.
▸Full profile — market position, technology, go-to-market, geography, history
Market position
As of summer 2024, FoldFlow-2 was characterized as the leading protein generative model, outperforming prior models such as RoseTTAFold Diffusion from the University of Washington; the company cites DeepMind's AlphaFold work in protein structure prediction as a reference point for the field.
The company positions its advantage in a multimodal generative approach that jointly models protein sequence and structure rather than structure alone, coupled with in-house wet-lab experimentation that feeds sparse empirical data back into its models, and in a team drawn from Mila combining generative modeling and protein expertise with access to Mila's research network and advisors including Yoshua Bengio.
Technology
Dreamfold develops generative models that reason jointly over protein sequence and 3D structure to address drug discovery problems. Its FoldFlow-2 model was built to exploit both 3D structural information and large protein sequence databases in a multimodal generative architecture, and is described as able to generate diverse, novel proteins unconditionally from random noise. Earlier described technology included active learning and GFlowNets applied to high-throughput biological experimentation for antibody design against membrane proteins. Continuous wet-lab testing supplies sparse experimental feedback that enriches the company's learned protein structure representations.
Go-to-market
Therapeutic applications in cancer, autoimmune disease and other severe conditions, with an emphasis on immune modulation.
Geography
Based in the Montreal ecosystem, with ties to Mila and participation in CDL-Montreal; backed by European investor OpenOcean.
History
The company was founded by Mila-affiliated researchers and participated in the Creative Destruction Lab's CDL-Montreal Artificial Intelligence stream in the 2022/23 cohort. It published research including spotlight presentations at ICLR and ICML, and by summer 2024 had developed FoldFlow-2, a multimodal generative protein model. It is a portfolio company of OpenOcean, with General Partner Tom Henriksson on the deal team.
Compiled by commissioned research from 8 cited public sources — announcements, filings, and press listed under research sources below.
Key figures
latest reportedCompany-reported or press-reported figures, each dated to when it was claimed — not independently audited.
Competitors · 4
by search overlap
Companies competing with Dreamfold for the same Google search keywords, organic and paid, via search-intersection analysis.
Timeline · 4
launches, deals, and filingsA funding news report stated Dreamfold raised $6,000,000, to be used for computational infrastructure and expanding its team of scientists and engineers.
$6M source ↗
OpenOcean lists Dreamfold as a portfolio company under its AI technology category, with General Partner Tom Henriksson on the deal team.
Dreamfold developed FoldFlow-2, a multimodal generative model that reasons over both protein 3D structure and sequence data; as of summer 2024 it was described as the leading protein generative model.
Dreamfold (legal name Wise Algorithms Inc.) took part in the Creative Destruction Lab CDL-Montreal Artificial Intelligence stream, 2022/23 cohort.
Dated company events from announcements, filings, and press; legal rows summarize public dockets and regulator releases.
In the news
▸Research sources · 8
primary sources listed
- Dreamfolddreamfold.ai · web
8 public sources were cited for this profile; the first-party ones are listed here.
Frequently asked questions
- What does Dreamfold do?
- Dreamfold builds generative AI models over protein sequence and structure to design protein therapeutics for cancer and immune disease.
- Who are Dreamfold's investors?
- Dreamfold's investors include OpenOcean, Panache Ventures, IQ Capital.