Microsoft's quantum program had to improve qubit reliability and speed while coordinating software, architecture, materials, fabrication, and measurement work across large, siloed datasets and globally distributed specialists. Manual experimentation and measurement cycles took weeks, complex parameter tuning was hard to scale, and hidden fabrication issues were difficult to detect quickly.
Reported impact
Microsoft accelerated quantum R&D for Majorana 2 by using agentic AI to automate measurement and analysis work, surface hidden fabrication issues, and help optimize materials and experiments, contributing to qubits reported as 1,000 times more reliable and supporting a revised target of achieving a scalable quantum computer by 2029.
Microsoft used Microsoft Discovery's agentic AI capabilities inside its quantum R&D workflow to support the Majorana 2 program. The platform lets the team deploy AI agents, guided by scientists, to reason over nearly two decades of heterogeneous project data spanning software, device architecture, materials, fabrication processes, and measurements. In practice, agents manage workflows, synthesize cross-disciplinary knowledge, correlate data trapped in silos, and recommend likely experimental targets before physical testing. The team also built specialized agents to automate qubit-state measurement, perform parallel voltage adjustments, and generate 3D maps of operating conditions that would be impractical for a single researcher. Additional agents filtered fabrication data and identified anomalies such as a miscalibrated temperature sensor. Outputs are served back to scientists in a human-in-the-loop process that informs materials choices, fabrication optimization, experiment design, and root-cause analysis for device reliability.
Historical quantum project data from software, architecture, materials, fabrication, and measurement systems is analyzed in Microsoft Discovery by specialized AI agents. The agents correlate siloed information, run simulations, automate measurement and voltage-adjustment loops, build 3D operating maps, and flag anomalies; their recommendations are then reviewed by scientists and fed back into device design, fabrication, and experiment execution.
'AI agents run the process automatically and continuously, building a 3D map of the conditions that a single scientist would never be able to do in the same way.'
'Majorana 2's new features include a new materials stack enabling a 1,000-fold improvement in reliability over the prior generation of qubits, with a mean qubit lifetime of 20 seconds and instances lasting as long as one minute.'
'Microsoft now expects to achieve a scalable quantum computer by 2029, cutting its original timeline in half.'
'The quantum project also has huge quantities of data-nearly two decades' worth, in many different formats. Before AI, the data was stuck in silos.'
'The quantum team's own scientists and engineers have been using the agentic AI capabilities in Microsoft Discovery to manage workflows, automate measurements, optimize fabrication, pinpoint previously unnoticed flaws and propose new solutions.'
'Using agentic capabilities available in Microsoft Discovery, the team was able to create an AI agent specialized for this job, which cut the cycle time by orders of magnitude.'
'the team developed an AI agent that was able to combine physics, device and institutional knowledge to filter raw data from the quantum team's fabrication process and sniff out an uncalibrated temperature sensor reading that was throwing things off.'
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Traceability
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
Related
In-house Sample Sep 1, 2026 Thomson Reuters General-purpose LLMs can produce subtle but consequential errors in legal and compliance work, where professionals need trustworthy, citation-backed analysis across very large document sets. As of Q1 2026, CoCounsel has crossed one million users CoCounsel · Practical Law · Tabular Analysis In-house Sample May 12, 2026 Cohesity Cohesity needed to reduce friction in enterprise support by replacing manual forms, portal navigation, and low-context case intake that slowed diagnosis and increased customer effort during business-critical incidents. approximately 50% of interactions resolved by customers themselves without ever needing to open a case Cohesity NetBackup · Entitlement management system (VEMS) · NextGen Support Portal In-house Sample Feb 5, 2026 LG AI Research LG AI Research needed to extend AI systems built for high-stakes manufacturing into financial market workflows where forecasting accuracy alone was insufficient; investors also needed narrative reasoning, governance,… 4-week horizon AI-Powered Equity Forecast Score · EXAONE · EXAONE Business Intelligence 