AMRx: How Innominds and SCIINV Use AI to Support UTI Care and Antibiotic Choices
The digital platform uses clinical information to provide early predictions about urinary tract infections and antibiotic resistance, supporting doctors while conventional investigations proceed.
Innominds and SCIINV Biosciences have developed AMRx, an artificial intelligence and machine learning tool designed to help doctors assess the likelihood of a urinary tract infection and consider antibiotic resistance patterns earlier in care.
Innominds announced the partnership in June 2024. A company update published on 4 September 2026 describes the platform as a clinical decision-support solution that works alongside clinical judgement and appropriate diagnostic testing.
The problem it addresses is the gap between a patient’s initial assessment and the availability of microbiology results. Earlier, reliable information could help clinicians make more informed decisions during that interval.
How AMRx uses clinical information
The software uses information such as patient history, presenting complaints and symptoms to generate predictions. Its published product description says the prediction process does not require a new sample or specialised testing device.
Its intended functions can be grouped into three areas:
| Function | Intended output |
|---|---|
| Infection assessment | An estimate of whether a urinary tract infection is likely |
| Organism-related prediction | An indication of possible Enterobacteriaceae involvement |
| Resistance assessment | Predictions about antibiotic susceptibility or resistance patterns |
Enterobacteriaceae is a family of bacteria relevant to urinary infections. Identifying a possible bacterial group through a model is a predictive step, rather than confirmation that a particular organism has been isolated from the patient.
The practical aim is to organise clinical information into outputs that a doctor can consider during an assessment.
Why earlier guidance matters
AMRx targets a point in care when clinicians may need to decide on initial treatment before complete laboratory results are available. Such initial prescribing is often described as empirical treatment: it is informed by the available clinical evidence while more specific information is pending.
The developers position the platform as an additional source of information for this stage. Its value would depend on whether the predictions help clinicians make appropriate decisions for the patients they are treating.
This sits within a wider public-health challenge. The World Health Organization’s July 2026 fact sheet states that approximately one in six laboratory-confirmed bacterial infections worldwide were resistant to antibiotics in 2023. It also identifies antimicrobial misuse and overuse as important drivers of resistance.
Improving antibiotic decisions therefore involves both treating infections effectively and avoiding unnecessary or unsuitable prescribing.
Culture-free predictions complement laboratory evidence
AMRx is described as culture-free and sample-free because its software generates predictions from clinical information. Those terms describe how the digital tool operates.
Laboratory culture and susceptibility testing answer different questions through the examination of patient samples. A model-generated resistance prediction is not a direct measurement of how bacteria from an individual patient respond to an antibiotic.
Innominds’ September 2026 explanation places the platform alongside conventional microbiology that is already in progress. This is an important part of its intended clinical role.
For healthcare providers, the useful workflow is one in which a clinician can review early predictions, continue appropriate investigations and reassess decisions as more evidence becomes available.
The research and institutional links behind the platform
AMRx brings together software engineering, biotechnology and academic research. Innominds contributes digital and product engineering expertise, while SCIINV Biosciences focuses on diagnostic approaches related to antimicrobial resistance.
A January 2025 University of Hyderabad announcement identifies SCIINV as a Hyderabad-based startup incubated at ASPIRE BioNEST. It also describes collaboration with the Sri Sathya Sai Institute of Higher Learning.
The BIRAC 3i product listing identifies that institute as the innovator and explains how the prediction models were developed. Clinical information was collected prospectively, with informed consent, from seven hospitals in southern India.
The listed inputs included symptoms, existing medical conditions and clinical history. This provides a clearer picture of the platform’s research foundations than a general description of AI alone.
Speed claims need to be assessed in clinical context
AMRx is presented as a way to provide information within minutes. A Microsoft customer profile describes a five-minute turnaround for the digital platform.
That is a published product claim about the software workflow. Its practical value depends on the completeness of the input information, the reliability of the predictions and how clinicians use the results.
A fast output becomes clinically useful when it supports better decisions. Relevant evaluation measures include missed infections, incorrect susceptibility predictions and performance across different patient groups.
WHO’s considerations for healthcare AI emphasise external validation, data quality and a clearly defined intended use. These principles are particularly relevant when a model’s output could influence antibiotic prescribing.
Technology supports delivery and integration
AMRx is listed on Microsoft Marketplace by Innominds Software. Microsoft’s customer profile identifies Azure Machine Learning, Azure SQL and Azure OpenAI among the technologies used in its development.
These services support the digital platform, but the choice of software infrastructure does not itself establish clinical effectiveness. That assessment depends on the model, its evidence and its performance in the intended setting.
For hospitals and clinics, adoption would also depend on practical details: how clinical information is entered, how clearly predictions are displayed and how the output fits into existing records and consultation routines.
A system that adds useful information without creating excessive administrative work is more likely to become part of everyday clinical practice.
Potential value in settings with limited laboratory access
For AMRx, an important proposed use is providing earlier decision support where laboratory capacity is constrained. The developers argue that a software-based approach could make relevant information more accessible without requiring specialised diagnostic equipment for the prediction itself.
This could be useful where results take longer to reach the treating clinician. The potential benefit is access to an additional assessment tool during the waiting period.
However, the software would still need dependable clinical inputs, suitable digital infrastructure and staff who understand its intended use. Its implementation would also need to preserve access to appropriate testing and follow-up.
The strongest evidence of value would come from routine clinical use showing more appropriate prescribing and meaningful patient benefits.
What will determine the next phase
AMRx illustrates how clinical information can be used to produce predictions at an earlier stage of care. The collaboration combines an identified medical need with software development and research experience from Indian institutions.
The next phase will be defined by the strength of clinical evaluation, performance across healthcare settings and the ease with which doctors can interpret and use the results.
For the developers, the central opportunity is to demonstrate that earlier predictions improve decisions in practice. For healthcare providers, the priority is a tool that contributes useful evidence while keeping patient assessment, diagnostic testing and clinical judgement connected.
For more insights navigate: PressPire | Stories That Shape the Future
Read previous article here : coal-india-1200-tpa-rare-earth-magnet-plant/
Follow PressPire on Facebook | Instagram | LinkedIn | YouTube
PressPire | Stories That Shape the Future