DATA. DRIVEN. DISCOVERY.

Ask a question and start a conversation with ARC, your AI Research Collaborator. You can ask for gene overviews, potential hypotheses, literature reviews, experimental protocols, and more. Think of ARC as a lab companion there to answer any of your questions and offer support throughout each step of the end-to-end research process.

With Heureka, you can access our data-driven hypotheses, chat with ARC (your AI Research Collaborator), explore our database of 20,000+ genes, metabolites, pathways, etc, access bioinformatics and interpretation tools, and much more.

At its core, Heureka is data-driven, not language driven. This means that our insights are based on empirical data associations and traceable to data artifacts without being subject to author bias, hallucinations, or dependencies on published literature. Each output has a data artifact tied to it so you have access to trusted insights with greater credibility and validity. Additionally, we put each output through a series of adversarial models, verification mechanisms, and multi-step tests to maximize the reliability of our outputs. We also have a suite of tools and agents custom-built for researchers to advance discovery. This approach sets us apart from traditional LLMs and science AI services that are subject to errors that can erode reliability.

Yes. We know that labs explore a wide range of topics and, similarly, researchers have a diverse set of interests and experience levels. That's why we've designed Heureka to be able to meet the needs of all researchers. Our knowledge network includes an expansive collection of data so you can get relevant insights across disciplines, start with broad overviews or dive deeper into nuanced topics, and even run advanced statistical analyses whether you have no computational background or you are an experienced bioinformatician. So, if you're a seasoned PI or just starting your research journey, Heureka can help you.

We've designed Heureka so that you can use our full set of tools regardless of your experience level with bioinformatics, statistics, coding, etc -- no technical expertise required. Every project, analysis, and interpretation is custom-built and you can ask follow up questions in ARC to get the most out of your report package.

Heureka generates a custom plan, statistical analysis, code, and interpretation for each project. We do not use templates or pipelines. Our bespoke approach maximizes the relevance of outputs to fit your data, project objectives, and research interests. Additionally, each part of our analysis goes through multiple rigorous tests, adversarial models, and checks to validate its validity and reliability so researchers can trust outputs.

Heureka does not train on your data. When you upload questions, ideas, or datasets, they are private to your instance. Our tools can process your data to give you unique insights based on your projects and those findings will stay private to your account. Similarly, when other accounts upload data to their private instances, you will not have access to their data and their insights will not impact outputs or results surfaced to you. Heureka upholds the highest security and data privacy standards to ensure that each user's data remains protected and only validated, credible insights from trusted sources are shared with researchers.

QuartzReport is Heureka's bioinformatics solution. You provide data and we run multi-omic statistical analyses as well as data-enriched interpretation to prepare a comprehensive, enhanced report. Each QuartzReport includes a report, custom code, data files, and publication-ready figures. Our analysis is custom-generated for every project to ensure relevance and completeness. You can also 'chat' with the report via ARC to dive deeper into specific topics, ask follow-up questions, design next steps, and more.

You can initiate a QuartzReport by uploading a dataset and chatting with ARC, Heureka's AI Research Collaborator. Simply request an analysis in the chat. The more detail you provide (e.g., project overview, research objectives, context, definitions and labels, etc), the better your report will be. After your analysis is complete, you will receive a QuartzReport package in your email -- note that this can take up to 24 hours and sends to the email registered to your account (check your spam folders if you don't see a message from QuartzReport within 24 hours). This package includes a full report with key findings, data-enriched interpretation, code files, data files, and high resolution, publication-ready figures. You can also initiate a QuartzReport via email. Simply attach your file (<25 MBs) to an email along with project details and a description in the message body (methods, definitions, labels, objectives, output requirements, etc). Then send it to data@quartzreport.com. You will receive an analysis plan for confirmation. Once your report is ready, you can hop back into ARC and 'chat' with your data report to ask follow-up questions and continue discovering.

We process public data through our proprietary AI/ML algorithms to develop data objects. We then compare insights against published scientific literature to determine if a hypothesis is "known" or "unknown." We continue to update our hypotheses and knowledge database so we serve the most up to date ideas to researchers using Heureka.

We source data from credible, validated public sources including PubMed, TCGA, and other trusted institutions. You can also upload your lab's data to your private instance to help you generate unique hypotheses, ideas, and insights specific to your research.

From our empirical data associations and data analysis, we generate data objects which we then classify as "known" or "unknown" in scientific literature. The underlying data objects that generate hypotheses follow the same methodology and process as those that are validated by literature to populate our gene summaries and topic overviews. Over time, our hypotheses are tested through experimentation and migrate to 'known' categories once proven.

Some data may appear in published literature missed by our classifier, so they will appear in the "hypothesis" section instead of the "known gene summary" section. For the gene summaries, all data objects have a corresponding publication, so this should limit the risk of any unproven facts being classified as "known." And remember: these are hypotheses after all.

We dynamically update our data and call papers and scientific literature in real-time, so you have access to the latest validated scientific knowledge.

Each researcher can cite Heureka as they deem appropriate -- we do not have a specific reference requirement. Generally, scientists note Heureka in the 'Acknowledgements' section of their papers.

Yes, you can try Heureka for free for 7 days. Our onboarding process will get you up and running -- and discovering -- in no time. We'll start you off with a credit balance that you can use to explore Heureka's tools and features for a week. After 7 days, your credits will expire, but you can add more to your account for a seamless onboarding experience.

That's right! We know how limited research budgets can be so we've designed a system that works for all sizes and types of labs. Instead of a subscription of license fee, researchers on Heureka's platform use credits. This enables labs to pay as they go and add members for free, increasing access to our tools.

Each account uses Heureka credits to track use. Chats, queries, and analyses each have variable computational complexity, so credits help measure utilization so labs can pay as they go without the need for subscriptions or licenses. When your credit balance runs low, you can always add more to your account and even get volume bonuses.

Because we provide all of the Heureka data and hypotheses from the first day, we cannot offer refunds.

When you open an account, you are the account's administrator. In the Heureka portal, click on 'Profile' and you should see a 'Team Members' tab. Here, you can invite new team members to join your account. Once they join, they will have access to your account credits, but, for data security and privacy, data won't be shared across users. Adding users is free and there is no limit to the number of members that can join your account.

There's no limit on the number of users that can join an account and no charge for adding members to your account. You can invite everyone in your lab, department, or organization to join your account. Data won't be shared between users, but everyone in the account will have access to the account's credits. As long as there are credits in the account, all approved members can keep discovering with Heureka!

Yes! We know researchers often have affiliations and collaborations across organizations. That's why individuals can be members of multiple accounts. First, have the account administrator add you to their account. When you log in, select the relevant account for that session. Your credits transactions will be recorded in that account's ledger. To switch between accounts, log out, sign in, and select the desired account. For security and privacy, your data will stay tied to your user profile and won't be shared across accounts.

At Heureka, we do not have recurring billing. When your account runs low on credits, you can add more as needed. Follow the payment link and the credits will be automatically added to your balance.

Yes, the payment portal only shows your base credit purchase -- bonus credits will be added to your account after the transaction. If you still think you should have more credits, feel free to reach out to us at support@heurekalabs.co.