Between Papers analyzes technical research across entire fields to uncover emerging capabilities, non-obvious commercial applications, and business opportunities — with every important conclusion traceable back to the evidence.
25–100 papers. One research question. One evidence-backed investigation.
A new electrolyte kept 93.8% of capacity well below freezing in lab pouch cells. Two commercial cells kept about 95% at 3C. One estimate puts factory energy for post-lithium cells below lithium-ion. But modeled cells hold less energy per litre than LFP, and commercial cells aged faster in the cold.
“…capacity at - 20, - 40, - 60, and - 70°C, respectively (Fig. 4b). The pouch cell with the EMTMSA-based electrolyte maintained high capacity retention, even under simultaneous low-temperature charging and discharging at 0.01 C, retaining 81.1%, 59.7%, and 41.9% of its RT capacity at - 20, - 40, and - 60°C, respectively (Supplementary Fig. 18).”Asymmetric sulfonamide design enabling high-voltage sodium-ion pouch cells in wide temperaturep. 7Electrochemical performance of SIB pouch cells at room and low/high temperatures
Three investigations, published in full. Every count below is read from the investigation it describes, and every citation in them opens the passage it rests on.
The answer beginsSodium-ion has moved from lab to factory, but the evidence does not show it making home or grid storage cheaper than lithium iron phosphate (LFP) yet.
The answer beginsOur judgement is that aircraft surveys and AI-screened satellites are closest to reliable, large-scale use.
The answer beginsThe most dependable result is machine-learning triage placed after a bank's existing alert rules.
Important technical shifts rarely arrive in one neat paper. The signal is distributed across experiments, disciplines and years of research.
Hard carbon has not reached its expected cost edge over graphite because feedstock is imported coconut shell and supply chains are immature.
C-027-0002C-018-0003Waste wood and treated switchgrass yield lab hard carbon with higher capacity than untreated or lower-grade versions, without strong acids.
C-021-0001C-021-0002C-017-0005C-017-0002Build regional hard-carbon plants that turn waste wood or energy grasses into anode material for European sodium-ion cell makers. The aim is to replace imported coconut-shell carbon and cut the anode cost that one model blames for sodium-ion packs costing more than LFP.
Why no single source says this. The cost study names feedstock as the bottleneck but not a fix; the materials papers show performance but no cost or supply case.
See it in the investigationScale is the smaller part of it: the batteries investigation started from 25 seed papers, looked at 396 further sources and read 6 of them in full to find connections like this one.
Four kinds of output, each shown here as the investigations actually wrote it.
Where independent evidence converges.
Two independent models agree that sodium-ion reaches parity with LFP only under conditions: supply shocks, high lithium prices, or better materials and designs.
Where findings conflict — or only appear to because experimental conditions differ.
In most aerial campaigns, a tiny share of sites produces 50%-81% of emissions. A national facility model instead puts 70% of US emissions below the 100 kg h level. One review finds that the 100 kg h threshold catches 95% of emissions, but only among point sources that aircraft had already detected. The populations differ, and neither result refutes the other.
What the combined research suggests engineers may now be able to do.
In blind controlled releases all three tested aircraft caught 100% of releases above 50 kg h, and one lidar finds 1.2 kg/h sources half the time. Size estimates for single releases vary widely; a simulation suggests matching an annual component survey needs about 3 flights a year.
Where those capabilities could create useful products, services, or businesses.
A mule-risk score for incoming Faster Payments accounts that combines account behaviour with network features and explains each flag, so the receiving firm can hold or freeze funds before they move on.
Discover emerging technologies, adjacent-domain solutions, and capabilities that could become your company's next product or service.
Explore product innovationUnderstand what research actually enables, what remains uncertain, and which commercial directions deserve further diligence.
Explore commercializationFind applications for your work — including possibilities outside your immediate discipline.
Explore research commercializationUpload 25–100 papers and tell Between Papers what you're trying to understand.
Technical findings, experimental conditions, limitations, agreements, and contradictions are extracted and connected across the corpus. Each claim is checked against the passage it came from.
Findings are connected into emerging technical capabilities, including potentially useful connections from adjacent fields and, where it helps, sources from outside your corpus.
Capabilities become potential applications and business opportunities, while evidence, assumptions, and uncertainty remain visible.
The whole method, with the counts behind it: how an investigation is made.
Between Papers does not ask you to trust an AI-generated conclusion. Inspect the evidence, assumptions, contradictions, and uncertainty behind it.
Important claims remain connected to their source material. Every claim is checked against the passage it cites before it is kept, and every number in a finding must appear in a kept claim.
Batteries investigation: 314 checks run against the source passages before 158 claims were kept.
Conflicting evidence is not silently averaged away. Every investigation has a section for where the sources disagree.
“The populations differ, and neither result refutes the other.”
Results under materially different experimental conditions are not automatically treated as equivalent.
“The difference is test conditions against field conditions. It is unresolved.”
The product distinguishes what the papers demonstrate from what it infers, and says so in the text.
“our inference is that satellite-only programs would miss much of the total”
Claims caught and corrected are counted, and sources that could not be fetched are listed, not hidden.
Never: a promise of accuracy. The chain is the promise.
What recent advances in sodium-ion and other low-cost batteries make cheaper home and grid energy storage commercially viable, and what products or businesses become possible?
The answer beginsSodium-ion has moved from lab to factory, but the evidence does not show it making home or grid storage cheaper than lithium iron phosphate (LFP) yet.
Presented as it was produced: the answer first, then the findings, the opportunities, the ideas from combining sources, and the evidence behind every one of them.
Research does not stop when your report is finished.
Monitor watches the landscape around your questions and tells you when something materially changes — a conclusion strengthened, a barrier removed, a transferable result from another field.
See what Monitor will doAsk one serious research question. Analyze a meaningful technical corpus. Receive the complete evidence-backed investigation.
11% less than buying singly
Three investigations, each one question and one complete report. They don't expire.
Offered when you unlock your first report, and from your account.
33% less than buying singly
Ten investigations, each one question and one complete report. They don't expire.
Offered when you unlock your first report, and from your account.
Start free: the agent reads your seed papers and gives you the answer and the first findings. Pay only to continue the same investigation to the complete report.
Not satisfied with a full report? A refund within 7 days of purchase — single investigations only; packs are final. The refund policy.
One question. One investigation. Buy one, or a pack.
No. Between Papers treats the corpus as a connected body of evidence. It looks for agreements, contradictions, dependencies, and combinations of findings that may imply something no individual paper states directly.
General-purpose models can read and summarize research. Between Papers imposes a structured investigation process designed specifically around cross-paper evidence, technical conditions, provenance, emerging capabilities, and commercial implications.
Yes. Important conclusions can be traced back through supporting findings to source passages.
No. Conflicting evidence, different experimental conditions, limitations, and uncertainty remain visible.
When useful, the investigation can use targeted external research to test assumptions, investigate adjacent fields, and assess commercial context.
No. Between Papers identifies evidence-backed possibilities and explicitly surfaces what remains unproven.
Read a real investigation first. Then run one on the question you actually need answered.
The flagship: cheaper batteries for homes and the grid, 25 seed papers, 396 sources looked at.