SciSpace AI’s New Research Agent Updates Explained
- Jo Clubb
- 1 hour ago
- 6 min read
In this post, I demonstrate the latest updates to AI research platform, SciSpace, including Zotero integration, library search, report writing, and notebooks.
AI research tools continue to evolve rapidly, and one of the platforms I have been using myself is SciSpace. I've previously discussed SciSpace as part of my 5 AI tools for Sports Scientists and now they've rolled out several major updates to their AI Agent, all aimed at reducing research friction and helping users spend more time thinking critically about the literature, rather than simply organising and collating it.
In this article, I walk through those new features step by step, from uploading and chatting with a single PDF file, to importing your library from Zotero, along with the new report writing tool to create a literature review within minutes. This demonstrates how SciSpace can support real-world research workflows for sports scientists, both researchers and applied practitioners.
Want to try SciSpace for yourself? Use the link and codes below for a discount on their paid plans 👇
JCLJAN20 : 20% off SciSpace Premium or Advanced monthly plans
JCLJAN40 : 40% off SciSpace Premium or Advanced annual plans
What Is SciSpace?
SciSpace is an all-in-one AI research platform designed for academics, researchers, and practitioners. It combines several AI-powered tools within a single ecosystem, including:
SciSpace Agent for interactive research queries
A dedicated BioMed Agent for biomedical research
Literature review and report writing tools
Citation generation
Scientific illustration generation
At the time of recording, SciSpace reports that its Agent connects to over 150 tools and 59 research databases, with access to more than 280 million academic papers. The most recent updates are focused specifically on improving how the Agent interacts with your own research library and how outputs are structured and stored.
In this walkthrough video, embedded below, I deliberately started with a near-empty SciSpace profile. My library and notebooks were essentially blank, which mirrors the experience of many first-time users. This made it possible to clearly demonstrate how the new features integrate into an existing workflow rather than requiring it to be rebuilt from scratch.
Analysing a Single Paper with SciSpace
Before exploring the more advanced features, it is worth understanding the basics.
SciSpace allows you to upload a PDF directly or import files from services such as Google Drive.
To demonstrate this, I uploaded a really interesting recent full-text publication from the British Journal of Sports Medicine - Qualitative and quantitative situational characteristics of muscle strains in sports: a systematic review and meta-analysis (Finnern et al., 2026) - and asked the Agent to summarise the paper.
The output provided a structured overview including:
Research question and study aim
Study design and methodology
Primary and secondary findings
Conclusions and practical implications
What is particularly useful here is the clarity of structure. Rather than a generic paragraph summary, the Agent organises information in a way that mirrors how many practitioners already read and evaluate research.
SciSpace also allows follow-up questions within the same chat, enabling deeper interrogation of a single paper. One example shown in the video was generating a scientific illustration based on the summary, which can be downloaded for further use. As with any AI-generated output, these visuals should always be checked carefully against the original paper.
Zotero Integration: Connecting Your Existing Research System
One of the most significant updates is the new integration between SciSpace and Zotero.
Zotero is widely used as a reference manager across academia and applied sport science (see my demonstration below), and one of the biggest friction points with AI tools has been the need to manually upload and re-upload PDFs that already live elsewhere.
With the Zotero integration, SciSpace can:
Connect directly to your Zotero account
Import existing folders and collections
Analyse those papers automatically
Populate your SciSpace library with structured metadata
Once connected, my existing Zotero folders appeared immediately inside SciSpace, including topic-specific collections such as my folder on women’s football research. This removes a major barrier to adoption and allows SciSpace to sit on top of an existing research system rather than replacing it.
SciSpace has also indicated that this is the first of several planned integrations, with tools such as Notion, OneDrive, Slack, and GitHub expected in the near future.
Library Search with the SciSpace Agent
Once papers are in your library, the SciSpace Agent can now query them directly. For example, I asked the Agent about ACL injury risk in women’s football, restricting the search to papers already in my library. The output clearly showed:
Which specific papers were used
A summary of key findings
Live links back to the original publications
This is a critical distinction compared to generic AI tools. You can see exactly where information is coming from and maintain transparency around sources, which is essential for applied decision-making and academic integrity, as I discussed at length in my chat with Professor Greg Haff on AI.
Deep Review Search Across the Literature
In addition to querying your own library, SciSpace also allows for broader literature searches across the available databases.
Using a more focused prompt on factors contributing to ACL injury in women’s football, the Agent performed a deep review search. This process involved:
Running multiple search queries
Screening and filtering thousands of papers
Narrowing results down to the most relevant studies
Producing a structured report with live citations
Within a short period of time, the Agent generated a comprehensive review based on a curated subset of highly relevant papers. It initially found 1750 papers on the topic, filtered that down to 335 'top papers', and added an additional 53 based on relevant references.
While this does not replace reading the literature, it can dramatically reduce the time spent on initial scoping and theme identification.
SciSpace then generated (within minutes) a report from 50 of the most highly relevant papers, broken down into the introduction, purpose and scope of the review, methodology, results, overall synthesis and conclusion, and references. Throughout the report, live links to the original papers are provided for deeper exploration.
Using Notebooks to Store and Revisit Your Work
Another key update is the ability to save outputs directly into SciSpace notebooks.
Reports, summaries, and analyses can be stored categorically, creating a clear audit trail of your thinking and allowing you to return to specific projects later. For longer-term research or applied projects, this is particularly useful for maintaining continuity and avoiding duplicated effort.
Again, this should be viewed as a starting point. The value lies in having a well-organised, structured foundation that you can then refine, critique, and build upon manually.
What SciSpace Does and Does Not Replace
It is important to be clear about where tools like SciSpace fit into the research process.
SciSpace does not replace:
Reading papers in full
Critical appraisal of methods and limitations
Domain expertise and contextual judgement
Responsible academic or applied writing
What it does replace is much of the manual friction at the front end of the process. It helps you move from a blank page to a structured overview more efficiently, freeing up time and cognitive energy for higher-level thinking.
Used appropriately, it functions best as a research assistant, not an automated decision-maker. For more on that, take a look at my discussion on AI with Prof Greg Haff.
Final Thoughts
SciSpace’s recent updates, particularly the Zotero integration, library search within the Agent, report-writing sub-agent, and notebook functionality, represent meaningful steps towards more integrated AI-supported research workflows.
For sports scientists working with large and complex bodies of literature, these tools can offer a more efficient starting point, provided they are used critically and responsibly.
If you are interested in exploring SciSpace yourself, you can use the link and codes below for a discount on their paid plans 👇
JCLJAN20 : 20% off SciSpace Premium or Advanced monthly plans
JCLJAN40 : 40% off SciSpace Premium or Advanced annual plans
FAQs: SciSpace and AI Research Tools
What is SciSpace used for?
SciSpace is an AI-powered research platform designed to support literature reviews, paper analysis, report writing, and citation management for academics, researchers, and practitioners.
Can SciSpace analyse my own research papers?
Yes. With the new Zotero integration and library search feature, SciSpace can integrate, analyse and summarise papers from your own curated library, not just papers found through general searches.
Does SciSpace replace reading academic papers?
No. SciSpace should be used as a starting point to structure and summarise research. Full-text reading and critical appraisal are still essential.
Is SciSpace suitable for sports science and applied practice?
Yes. While SciSpace is not sport-specific, its ability to manage large bodies of literature and generate structured summaries makes it particularly useful for applied sports science, medicine, and performance research.
What is the SciSpace BioMed Agent?
The BioMed Agent is a specialised AI tool within SciSpace designed for biomedical research, supporting more advanced queries related to clinical and biological topics.
Can SciSpace help with literature reviews?
SciSpace can assist with scoping, summarising, and structuring literature reviews, including generating reports with citations. These outputs should always be reviewed and refined by the user.