AI in Research Administration: What Lyn Does

By using AI in research administration, offices dealing with sponsored programs can handle their daily workloads efficiently, even as they continue leveraging human judgment for critical compliance decisions. Tools like Lyn, which is the embedded AI assistant inside Streamlyne Research, are extremely necessary for administrators in modern research offices, whose roles typically include managing tight submission deadlines and agreement reviews. The solution streamlines administrative labor, particularly by drafting proposal budgets, while redlining incoming contracts against campus policies. In doing so, the AI also searches policy libraries effectively and answers data queries, making it easier for human professionals to evaluate risks and approve every final action.

What Is Lyn in Streamlyne Research?

Efforts to understand how Lyn works should take into account a notable boundary: Lyn is not the entire eRA platform.

As a product, Streamlyne Research is the core software system whose design houses distinct operational modules, including Proposals, Awards, IRB (Human Subjects), IACUC (Animal Welfare), and Institutional Reporting. On the other hand, Lyn is simply the AI assistant built inside Streamlyne Research, specifically with the goal of helping administrators process and draft work faster across those modules.

Furthermore, it is worth noting that as a solution, Lyn operates under two strict security and operational boundaries:

Role-Based Access: Based on its design, Lyn allows the currently signed-in user to only access data that they are authorized to view within Streamlyne Research, thereby guaranteeing proper security.
Human Control: Despite its immense capabilities, Lyn does not save changes, submit forms, or finalize contracts automatically. Instead, the AI allows a human administrator to review and confirm every suggestion before any data is recorded.

How Does Using AI in Research Administration Support Budget Drafting?

The process of building proposal budgets manually can be time-consuming, especially if you have to rely on faculty notes, spreadsheets, and narrative justification documents. In any case, faculty often provide preliminary numbers in varied formats, necessitating the need for administrators to translate raw narrative notes into line-item budget structures. Lyn can help offices handle such demands efficiently. Remarkably, the AI simplifies this workflow by reading uploaded budget justifications in different formats, including PDF, Excel, or CSV, while supplementing the information with instructions that the user provides in plain English.

As a sponsored programs officer, you might receive several budget justifications from principal investigators, particularly during a typical pre-award deadline week. The traditional approach would be to key in the information manually, such as the personnel costs, equipment estimates, fringe benefits, and indirect cost rates. While this strategy is doable, it is highly time-consuming and inefficient. Lyn allows the administrator to overcome such challenges as the user can simply upload the justification file into Streamlyne Research and add relevant notes. Subsequently, the AI will read the file, while parsing the numbers and constructing the structured line-item budget seamlessly. The administrator then inspects every line, even as they adjust figures where necessary and confirm the budget. This design is intended to ensure that nothing on the record is saved until the user approves it, guaranteeing proper human oversight over the tool’s assistance.

As an administrator, you might find yourself dealing with contract negotiations that stall awards, sometimes even extending to weeks of delays. This issue often arises when the administrator has to line-edit complex legal language against university guidelines, while also considering past sponsor precedents and relevant state regulations. Lyn can help you overcome this challenge effectively, particularly by accelerating agreement reviews. In doing so, the tool scans sponsor contracts that the user uploads, comparing the file against the institution’s established negotiation rules, while also taking into account the sponsor’s negotiation histories and pertinent policy documents.

As Lyn evaluates an incoming agreement, it presents structured feedback for each flagged clause, making it easier for the user to identify potential issues. This feedback includes:

The original clause, which is basically the exact language that the sponsor used in their draft.
The suggested replacement language to align with institutional standards.
The reasoning, explaining to the user why the original clause could potentially pose a risk or violate policy.
The policy citation indicating the specific institutional rules backing up the suggestion.

Hence, an officer could rely on our AI solution to manage the workflow once a corporate or federal sub-award agreement arrives in the office, specifically via Streamlyne Negotiations. Rather than spending hours reading fine print line by line, the user could simply run Lyn’s agreement analysis, allowing the tool to make the process more efficient. The officer can then evaluate each recommendation, even as they accept or reject the edits before exporting a clean Word document with tracked changes ready to be sent back to the sponsor.

How Can I Use Lyn to Search Data and Policy Libraries?

Research offices frequently handle urgent requests for data updates, as well as for clarifications regarding policies. Lyn allows the user to obtain specific answers, particularly by serving as an instant search and query engine for campus data and policy libraries. A user can open a chat window within Streamlyne Research, which would enable them to ask direct questions about their institutional data using natural language.

Examples of such questions include:

Which proposals are due in September?
Which negotiations are open with the NSF?
Show me all pending sub-awards of above $200,000 for the Chemistry department.

Upon receiving such queries, Lyn retrieves the relevant records, presenting the answer directly in clear data tables and visual charts. This design is meant to help administrators access immediate status updates without necessarily building complex database reports.

Equally important, institutions can upload their official policies, along with negotiation guidelines and other pertinent rules directly into Streamlyne. This arrangement allows Lyn to index the library, especially in a way that makes it easier for administrators to access instant search capabilities.

Why Is Human Oversight Crucial When Using AI in Grants?

Even though AI tools like Lyn offer speed and automation, institutions still have to ensure and optimize regulatory compliance. Notably, federal sponsors and other pertinent bodies require universities to maintain strict accountability for all financial reports, as well as for grant submissions and contract terms. Therefore, while you can use Lyn to handle the routine administrative tasks like extracting data and cross-referencing text, human professionals make the final compliance and legal decisions.

Federal Guidance and Resources

You can learn more about the rules for federal grants and how to comply with research standards in higher education by consulting these resources:

National Institutes of Health (NIH): By exploring the NIH Grants & Funding Policy, you could familiarize yourself with how to prepare budgets appropriately while learning more about allowable costs and the relevant rules.
National Science Foundation (NSF): Reviewing the NSF Proposal & Award Policies & Procedures Guide (PAPPG) could help you stay updated about the current standards for preparing proposals.

Frequently Asked Questions

Can I purchase Lyn separately?

No. Lyn is not a standalone product. Rather, it is an AI assistant that is embedded directly into Streamlyne Research, where it operates within the ecosystem, providing key capabilities for administrators.

No. Lyn’s design is meant to allow it to adhere to strict role-based access controls and institutional boundaries. The tool only searches data within your institution, while users can only view data that their specific permissions allow.

Does Lyn submit budgets or send redlined contracts to sponsors automatically?

No. Lyn never saves data or sends documents automatically. A human user must review every report and edit, explicitly approving or rejecting the changes before they are saved to the system of record.

Author bio: Bramwel Saisi is an author and researcher specializing in federal grant compliance, institutional policy, and sponsored project operations.