naiio.org currently presents a “Closed site preview” page. It asks for a password provided to the association’s founders and experts, and says access is for internal discussion; the materials are not yet a public version of the Association’s documents. NPNT’s portfolio describes NAIIO as a platform for the National Association of Researchers and Research Organizations, built on professional social-network principles. Its stated stage is “In development and testing”; the portfolio says a closed version has been prepared in Russian, English, Chinese and Hindi. Because the site is password-protected, profile and network functions cannot be checked without authorized access. A restricted preview allows participants to discuss materials before public presentation. It does not by itself confirm that the Association has launched, or verify membership or accreditation of people and organizations. The platform’s practical value, profile-verification quality and ability to support professional collaboration require separate evidence; these outcomes cannot be inferred from the public access page.
The Two Paths website presents an 18+ age confirmation before its materials. Its notice says the site contains educational content about dependencies, illness, violence and other dangerous life situations. NPNT’s portfolio describes the project as a story about two possible life scenarios—with or without tobacco—and the possibility of changing course. It is a published social and educational project; the public landing page currently shows an age screen before continuing. This format may provide a prompt for discussing the consequences of choices and available support. It is not medical advice, does not assess an individual’s condition and does not promise to change behaviour. The project also does not replace a doctor or specialist support for stopping tobacco use. The accessibility and usefulness of the materials after age confirmation, how different audiences understand them, and any effect on visitor decisions require separate evaluation. These outcomes cannot be established from the public landing page alone.
AI / Science presents a conceptual demo for research: a route from a hypothesis to expert review. Demo text is not sent to a server. The public page shows four steps: an evidence map, design options for an observational study, sample-size scenarios, and a document package. It explicitly says this is an interface demonstration, not a scientific recommendation. A separate preliminary-review flow for ethics committee materials describes checks for completeness, consistency, ethical questions, and explainable comments linked to document passages. The concept aims to connect research preparation in a clear route and make comments traceable for researchers and reviewers. AI can suggest options and surface questions; decisions about design, ethics, and clinical appropriateness remain with qualified specialists and the committee. The public version is a conceptual demo environment. The page does not establish deployment, clinical validation, or achieved performance. The accuracy of suggestions, quality of sources, and any effect on timelines or research quality require separate evaluation in a real process.
Not Instead is an educational catalogue about folk and complementary methods used in cancer care. Its published methodology describes six steps for examining claims; the pages show an update date of 20 August 2026. The process starts by stating the exact claim and intended goal: treating a tumour, supporting symptoms, quality of life and prevention require different evidence. It then proposes looking for clinical guidelines, systematic reviews and human studies; checking safety and interactions separately; verifying the status of a product or method; and publishing limitations. Laboratory findings are labelled as hypotheses, not established effects in patients. The practical purpose is to help readers turn promotional language into a question they can discuss with their care team. The catalogue explicitly says it is not a ranking: a method’s presence is not a recommendation or endorsement. The project is educational; it does not prescribe treatment or determine whether a method is suitable for an individual. Medical decisions should be discussed with the treating professional. This information does not replace medical advice.
AI·SPB publishes a concept for an experimental legal regime covering narrowly defined medical AI scenarios. Its page identifies working version 0.3, dated 19 August 2026, and states that the EPR has not launched and is not an official project of the Government of Saint Petersburg. The concept lists ten distinct clinical candidates and a 48-month evaluation path across stages A0–A4, from an assistive prompt to a potentially legally significant action. These are scenarios for assessment, not a list of autonomous services already permitted. Each use would separately require evidence, a registered product, an operator, a clinical site and a legal pathway. Uncertainty is routed to a human. The practical value is a structured discussion of permission by task, system version and control conditions, with refusal, escalation, measurable outcomes and stop criteria defined in advance. The publication describes a concept; it does not itself authorize autonomous action. Reduced time, less harm or improved outcomes are aims to evaluate, not confirmed effects. Every scenario needs its own legal, clinical and technical assessment.
The CGA Online workspace structures a professional comprehensive geriatric assessment. Its public page lists 16 sections, more than 100 fields and an estimated 35–50 minutes. The workflow covers a person’s context and goals, functional and physical measures, red flags, a problem list and a team plan. Answers remain in the open tab’s memory, are not sent to the operator and are not stored in localStorage; closing or refreshing the page removes entered information. The confirmed status is a professional protocol, version 1.0. The methodology states that the tool has not yet undergone independent validation and is intended to support documentation and clinical discussion until prospective evaluation. Its practical value is to bring observations into one working structure, reveal incomplete areas and prepare multidisciplinary discussion with named responsibilities and reassessment. The service does not diagnose, prescribe or discontinue medication, or replace examination or the clinical team’s decision. Completeness, agreement between professionals, usability and effects on outcomes that matter to patients require separate studies; this version alone does not establish clinical benefit.
The public version of Wearables presents a research workflow for remote observation during outpatient anticancer treatment. It combines electronic patient-reported symptoms (ePRO), activity, sleep and heart-rate trends, measurement quality and pre-agreed signal rules. A change is sent to the clinical team for human review; an individual reading does not become a diagnosis or prescription. The confirmed stage is version 0.1, a research prototype using synthetic data. Participant enrolment, device connection and collection of real patient data are not open. Specific devices, measures, thresholds, observation windows, roles and endpoints must be defined in a separate protocol. The practical value of the concept is to make change between visits visible and leave an auditable trail: baseline, signal rule, priority, specialist decision and action outcome. The project does not change treatment, replace an oncologist or emergency care, or provide treatment recommendations to a patient. Data completeness, monitoring acceptability, false-signal frequency and any effect on clinical outcomes have not been established for this prototype and require its own study.
The public Empalife site presents its output not as a long recommendation list but as a plan with an evidence passport. Every action should answer four questions: what to do, why, how to measure the result and where to read the primary source. The route has five steps: choose one leading task, add only context that changes the decision, receive up to three actions, inspect the evidence and record the result. The confirmed stage is a closed demonstration prototype and technical pilot v0.6 dated 20 August 2026. Only synthetic or fully de-identified test files are permitted in the cabinet; real patient information must not be uploaded. The practical value is to reduce overload and separate a low-risk action from a question requiring a professional. Supplements and vitamins are assessed by indication, clinical guidance, systematic reviews, interaction risks and whether the result would change a decision, not by popularity. The service does not diagnose, prescribe treatment, change doses, promise longer life or replace emergency care. User benefit, route safety and the completeness of the evidence passport remain unconfirmed and require separate evaluation.
Trialeco has published a dedicated page describing how a system should be checked before a pilot. It sets out eight control areas: purpose and status; roles and human decisions; quality verification; data provenance and minimisation; access and action history; change and resilience; deviation handling; and regulatory and scientific foundations. The confirmed stage is a publicly described project framework and a demonstration using synthetic data. The site distinguishes “designed”, “shown in a demo”, “verified in a pilot” and “implemented”; compliance with each requirement must be confirmed for each configuration. The practical value is to define permitted data, role rights, a reference standard, critical errors, acceptance criteria, stop conditions and change procedures in advance. A reference to a regulatory document does not establish compliance automatically. The demo does not replace technical, clinical, operational or regulatory verification. The system may assemble deviation materials, but classification, CAPA, closure and effectiveness checks are approved by designated quality functions. Clinical, ethical and regulatory responsibility is not transferred to AI.
Empathic Balance has published an evidence map of cosmetic care for sensitive skin covering January 2016 to 11 August 2026. It includes 12 key publications; PubMed/MEDLINE is the primary source, supplemented by the Cochrane Library, clinical guidance and official databases. Market context is cross-checked against a peer-reviewed analysis of 88 products from 19 international brands. The confirmed stage is a public review updated on 11 August 2026. It groups solutions into four areas and defines five development principles. Its practical value is to turn literature and INCI analysis into questions for a formula: what to measure, which risks to consider and where direct confirmation is required. A paper about an ingredient creates a hypothesis but does not prove a product’s properties. The review does not confirm the efficacy of Empathic Balance products and is not medical advice. Final INCI lists, properties, age guidance and use instructions should be published only after finished products are assessed and registration procedures are completed.
Autowave openly describes how it checks publications about biological waves. The route has seven blocks: defining the term, measurement, study design, causality, reproduction, the boundary of transfer and source checks through databases and identifiers. The public methodology is version 1.1; the page states that its literature base and criteria were updated on 20 August 2026. The confirmed stage is a working research portal with a separate methodology, curated library and labels for “established”, “under study” and “hypothesis”. Its practical value is to give researchers and science editors a short verification route before an interesting observation becomes a strong conclusion. A spectral peak does not prove message transmission, an in-vitro effect does not prove an effect in people, and a published paper does not prove independent reproduction. The portal does not perform experimental validation itself or confirm clinical efficacy. Review completeness, the applicability of the criteria to individual fields and the usability of the route still require expert assessment and further development.
Cognitive Gym publishes six free browser exercises: counter-sequencing, colour against word, a Schulte table, Go / No-Go, spatial sequence and rule switching. No account or subscription is required. The site states that completed rounds, time and game metrics remain only in browser storage and can be deleted by the user; results are not sent to analytics. The confirmed stage is a published multilingual project with six working exercises, a results log, methodology and interpretation guidance. Sessions are designed to be short and support comparison only with a person's own performance on the same task under similar conditions. The practical value is a clear format for a mental warm-up and observing one's pace and accuracy without rankings. A metric describes a specific game task, not intelligence or health. Improvement within an exercise does not prove transfer to study, work or daily life. The project does not diagnose cognitive impairment, treat stress or promise dementia prevention. The optimal practice frequency and long-term effect of these versions remain hypotheses for a separate study.
The public Kite + AI site shows how its riding-conditions verification loop is structured. For Saint Petersburg and the Leningrad region, the site lists 10 physical stations and 11 observation channels connected from open sources. Source availability and freshness are checked every minute, while the planner compares observations with three forecast models, shoreline orientation and spot parameters. The confirmed stage is a published planner for Saint Petersburg and Hurghada, a spot catalogue covering 12 countries, and an interface where riders enter location, weight, experience, kite size and travel period. The portal also provides cameras, learning materials and wind-checking channels. Its practical purpose is to reduce manual comparison of scattered forecasts before a trip. The minute-by-minute check monitors data availability; it does not automatically retrain a neural network. A recommendation does not guarantee safety or replace in-person instruction. Before launching, riders must independently check actual wind, direction relative to shore, cameras, sensors, obstacles, rescue infrastructure, equipment and their own condition. Recommendation accuracy and any effect on safety remain hypotheses until independently measured.
Medicine Horizon opens the research landscape through three clear filters: therapeutic area, geography of research sites and the phase at which a hypothesis is being tested. Its public interface groups active ClinicalTrials.gov records while preserving the NCT identifier, English title and a direct link to the primary protocol. The sample shows recently published records; the complete dataset remains in the official registry. The confirmed stage is a published analytical project with views by condition, country and phase, together with methodology, a knowledge base and editorial reviews. The service checks the source version marker and explains how each view should be read. Its practical value is to help an R&D team identify questions for examination and then move to the source record, endpoints and study design. Study counts do not establish evidence quality, a phase does not promise success, and an algorithmic tag is only the start of verification. The project is not an exhaustive systematic review, a forecast of regulatory approval or medical advice. Classification completeness, decision usefulness and detected trends remain hypotheses until experts test them for a specific task.
The public TrialCost sign-in page makes an important project boundary visible: it does not expose calculations or research-centre data. Access uses a temporary password; after sign-in, individual workspaces and institutional information are separated by distinct accounts. The page is excluded from search indexing and public caching. In the NPNT portfolio, TrialCost is described as a platform for preliminary clinical-study cost calculation, from the structure of visits and procedures to a budget model for discussion. The confirmed stage is a protected first-version workspace and expert testing of the calculation logic. The practical purpose is to organise cost items into a traceable structure before financial negotiations and make initial assumptions visible to the study-start-up team. TrialCost does not approve a final budget and does not replace a contract, protocol requirements, financial agreement, or decisions by a sponsor and research centre. Calculation accuracy, reference-data completeness, time savings and applicability across protocols remain hypotheses until authorised specialists test them on real scenarios.
The public LEC AI site shows how submission, completeness review, expert assessment, meetings, follow-up and archiving can be connected in one digital process. In the demonstration route, document versions do not overwrite one another, material actions receive an author and timestamp, and conflicts of interest, quorum and deadlines become visible before a meeting. Every public example is labelled synthetic. The confirmed stage is an informational demo and a limited closed server pilot. The site explicitly states that the system has not been commissioned as a production personal-data system and is not intended for medical records, patient source data or real research documents. The practical purpose is to reduce the risk of losing versions, deadlines and responsibility across email and folders while preserving traceability. The platform may issue reminders or block an unsafe route, but authorised committee members retain ethical assessment and decision-making. Each committee configures its own rules and deadlines. Time savings, error reduction and usability remain hypotheses until independently piloted and measured.
The public Empalab site presents an organisational model for supporting cosmetic care during and after treatment. Instead of a universal “allowed or prohibited” list, the interface offers three routes: a lower-trauma option under a protocol, discussion with a clinician, or temporary postponement of a procedure. The prototype structure includes a safety passport, procedure review, a personal care plan, the Emma digital navigator, transfer of a question to a coordinator and post-procedure monitoring. The page labels the current stage as a prototype being developed with clinical experts. Its public interface models 24 protocols in an initial knowledge base, three routing levels and 30 days of monitoring, but these figures describe service architecture, not published clinical results. The practical aim is to help a person structure a question and recognise when clinician review is needed. Empalab does not replace the treating clinician and is not an emergency service; individual medical decisions remain with a clinician. Route safety, usability and impact on real outcomes require independent validation.
The public PROMCYCLE AI site describes not an off-the-shelf platform but four testable stages: an initial diagnostic, a task passport, a line audit and a shadow Proof of Value. A pilot begins with one production loss—such as downtime, lower yield, or excess raw-material or energy use—then defines the baseline, available signals, a specialist action and a stop/go criterion. The confirmed project stage is readiness for industry interviews and selection of a pilot line. No confidential process data are required for the first diagnostic. The practical purpose is to connect a model warning to a verifiable production action and an economic KPI. In shadow mode, AI does not control the process: technologists, quality staff and engineers review its output. Published scenarios are reference architectures, not confirmed client results. Economic impact, warning accuracy and transferability to a specific line remain hypotheses until a separate pilot on new data.
The NPNT portfolio presents Empalife as a research direction for personal navigation across health and life goals. The public version is labelled a closed demonstration prototype and technical pilot v0.6 dated 20 August 2026. It describes six functional tasks: concentration, physical fitness, sleep, stress, long-term health and a combined route. The method starts with one leading goal and limits the result to three actions with sources, constraints and a success check. The confirmed stage is demonstration of a product model using synthetic data. Only synthetic or fully de-identified test files are permitted in the technical pilot; real patient information must not be uploaded. The practical purpose is to reduce recommendation overload and show when a low-risk action may be enough and when professional review is needed. The service does not diagnose, prescribe treatment, change doses or promise a longer life. Medical decisions remain with a clinician. User benefit, module safety and suitability for real data remain hypotheses requiring separate validation and authorisation.
The NPNT portfolio presents AI·SPB as a published concept for an experimental legal regime for medical artificial intelligence. Public version 0.3 describes ten distinct clinical candidates, an A0–A4 autonomy scale and a proposed 48-month staged validation path. Each narrow task would require a predefined population, inputs, output, system version, refusal option, escalation rules and stop conditions. The confirmed stage is professional discussion of the concept. The site explicitly states that the regime has not launched and that AI·SPB is not an official project of the Government of Saint Petersburg. The scenarios are not authorised autonomous medical services. The practical purpose is to create shared terminology for distinguishing AI assistance, output substantively confirmed by a professional and autonomy that could be considered only after separate validation and legal authorisation. Clinical decisions, diagnosis, treatment and legally significant actions are not transferred to the system automatically. Safety, economic effect and the acceptable level of autonomy remain testable hypotheses; every scenario would require evidence, a registered product, an operator, a clinical site and a legal gateway.
The NPNT portfolio presents MIKA.help as an operating information and navigation service for patients and families. The project has worked in oncology since 2016 and is now expanding clinical-study search to other conditions in stages. Its public site offers ten disease areas, a quick search for Russia, a detailed questionnaire, post-diagnosis materials and compulsory-insurance and high-technology-care routes for oncology patients. The service compares information entered by a user with open registries, explains a possible match and preserves the primary-source link, recruitment status and available official contacts. The open questionnaire does not require medical-document uploads; an unknown answer can be skipped. The practical aim is to reduce the information gap between a complex registry record and the patient's next understandable step. A study returned by search remains an option to verify, not confirmation of participation. The service does not diagnose, prescribe treatment or determine eligibility. Actual recruitment, medical assessment and possible inclusion are confirmed by a clinician and the study site. Broader coverage and intelligent assistants remain development directions, not demonstrated clinical outcomes.
The NPNT portfolio presents Trialeco, a modular AI platform for clinical studies. Its public site describes six connected components: Protocol Guard for protocol review, Recruit for pre-screening, Comms for participant communication, Site Status for assessing site readiness, an optional Orchestrator, and a Quality layer. Trialeco Core connects data, roles and the history of actions. The confirmed stage is architecture review and preparation of limited pilots. The team is assembling reference sets, while the open demonstration route uses synthetic data only. Accuracy, speed, completeness and usability are intended to be measured separately for each module. The practical approach is to begin with one process, baseline measure, reference standard, accountable roles and stop conditions rather than the whole platform at once. Any expected process improvement remains a project hypothesis until tested in a pilot. Trialeco does not make decisions about a person: medical assessment, confirmation of protocol fit and study inclusion remain with an authorised professional. The public demo is not an operational medical system and must not contain real patient data.
The NPNT portfolio presents NAIIO, a closed environment for the platform of the National Association of Investigators and Research Organisations. The concept follows a professional social-network model for researchers, research sites, scientific organisations, early-career professionals and sector partners. Russian, English, Chinese and Hindi versions are included in the prepared environment. The confirmed stage is development and testing. The public address opens only a protected-preview page: a password is provided to founders and experts, and access remains available on the device for seven days. The full platform is not currently presented as open for unrestricted registration. Its practical aim is to create a multilingual professional environment where researchers and organisations can present capabilities and find potential areas for collaboration. Scenario review with founders and experts, further development of professional profiles and Arabic localisation remain planned steps, not confirmed results. A platform profile must not automatically be treated as proof of qualification, membership, organisational accreditation or permission to participate in a particular study.
Medical AI Agents is a research laboratory for narrowly scoped digital assistants serving patients, clinicians, researchers and coordinators. Its public structure describes four roles and four baseline routes: clinical-trial navigation, compulsory-insurance care, high-technology medical care and paid services. Each scenario includes sources, an action log, stop rules and transfer to a professional. The confirmed stage is a closed synthetic prototype using fictional cases, a source library and an action log, with no real medical data. The public site includes a market atlas, methodology, three analytical articles and Russian, English and Chinese versions. The approach is intended to test one repeatable task and measure errors before integration into a medical process. The project does not claim clinical efficacy, medical-device registration or deployment in a medical information system. An agent does not diagnose, prescribe treatment, decide study inclusion or replace emergency services; an authorised professional retains the clinical or research decision.
Autowave is a public research and science-communication atlas about how living matter transmits signals. It connects five scales, from molecular oscillations and ionic waves to cellular, tissue and organ-level dynamics. Nine thematic sections, a curated library, glossary, methodology, history of scientific ideas and editorial reviews are available in Russian, English, Chinese and Hindi. The confirmed stage is a published multilingual portal that accepts proposals for publications, questions and expert reviews. Its practical value is to give researchers a common language across biophysics, cell signalling, mechanobiology, spectroscopy and medicine. The editorial model labels established mechanisms, active research and hypotheses separately. AI supports literature search, spatiotemporal-data analysis and model checking, but does not replace experiments or independent replication. The portal does not establish clinical efficacy, provide medical services, diagnose conditions or select treatment.
Empathic Balance is a cosmetics project for sensitive, dry and reactive skin. Its public page describes gentle cleansing, daily moisturising, skin-barrier support and more intensive care formats. The confirmed stage is launch preparation: sales are still being prepared and testing is available by preliminary application. The first-stage form does not request diagnoses, therapy details or other medical data. The practical aim is to make cosmetic-care choices easier to understand. Published materials do not yet confirm testing outcomes, clinical efficacy, release dates or regulatory approval. Empathic Balance is cosmetics, not treatment, and the online selector is not medical advice. Skin disease and treatment changes require consultation with a qualified clinician.
The first edition of GENES ≠ FATE is now available in the NPNT portfolio. It is an independent educational atlas of genetic tests and evidence. Instead of starting with a gene name or a ready-made commercial package, the project starts with a medical question: which risk is being assessed, what the laboratory actually measures and whether the result can change a decision. The atlas currently contains 35 reviewed relationships and two language versions. Its materials distinguish causal variants, clinical modifiers, ordinary group associations and research signals. Dedicated sections cover inherited risk, pharmacogenetics, oncology, laboratory checks and genomic-data privacy. The practical aim is to help people and professionals separate a clinically actionable result from a probabilistic trait or a marketing promise. The site does not sell tests or accept DNA or medical-file uploads. It remains an educational reference: it does not diagnose, prescribe treatment or turn a statistical association into a personal forecast. Medical decisions require clinical confirmation and discussion with a qualified clinician.
Experts are reviewing a digital workflow for ethics review: document versions, deadlines, meetings, conflicts of interest, reports and process audit. The test covers terminology, completeness and usability; the system does not make an ethics decision on behalf of a committee.
The platform structures the cost calculation for running a clinical study at a medical or research centre and supports the economic preparation of study start-up. Expert testing covers cost items, calculation logic and usability; authorised centre staff still approve the final budget.
Together with kite-community representatives, the project launched its first integration of AI-assisted forecast analysis, shoreline directions, weather sensors and cameras. It helps compare conditions with a spot and kite size, but does not guarantee safety; the rider remains responsible for the decision and local conditions.
Clinical-research and R&D experts are reviewing filters, classification and the clarity of analytics built from open ClinicalTrials.gov data. The goal is to improve question framing and planning; detected trends remain navigational estimates, not guaranteed forecasts of medicine's future.
An interactive comprehensive geriatric assessment questionnaire has been created with AI support for navigation and explanation of functional domains. Experts will review scales, question order, result clarity and boundaries. It is a professional-tool prototype: it does not diagnose, prescribe treatment or replace a clinician-led assessment.
NPNT is forming a working group on an experimental legal regime for medical AI
Medical, technology and legal expertise is being assembled to discuss bounded pilots, accountability, observability and stop criteria. This does not mean the regime has been approved.
NPNT update · partner roles are disclosed after confirmation
Work in progress
NPNT representatives are contributing to discussions on a cross-sector digital project for people with cancer
The discussion concerns a portal that could connect traceable information, communication, and routes to consultation and research services. Scope and participants remain under development.
NPNT update · partner roles are disclosed after confirmation
Work in progress
NPNT is developing an industry-partner role in a project using Yandex Cloud
The project scope, responsibilities and pilot architecture are being clarified. Mentioning Yandex Cloud describes the intended technology stack and does not imply an official NPNT–Yandex partnership without separate confirmation.
NPNT update · partner roles are disclosed after confirmation
NPNT
NPNT develops an international platform for medical and industrial AI
The corporate site is gaining nine native language versions, including the languages of EAEU member states, an open project portfolio, a science library and a clear partner route.
NPNT news
New project
PROMCYCLE AI: from production data to a verifiable decision
The portfolio now includes a platform with biotech, pharma, cosmetics and food-production cases. Each scenario links data, an AI signal, a KPI and a specialist decision.