How to Build an AI Radiology Information System for Healthcare Enterprises?

When was the last time your radiology department completed a full day’s workload without delays or backlogs?

And how often do your imaging workflows operate without disruptions from scheduling errors, reporting delays, or compliance concerns?

If those feel like loaded questions, you are not alone. Enterprises across U.S. healthcare systems are waking up to a “quiet crisis” in imaging operations that legacy RIS platforms simply cannot solve. That is precisely why AI radiology information system development is moving from “optional innovation” to a strategic imperative.

Consider this:

• According to PR Newswire, the global AI in healthcare market is projected to surge from about USD 14.92 billion in 2024 to USD 110.61 billion by 2030, growing at a healthy CAGR of 38.6%.

• Reports by Grand View Research, In the U.S. alone, the AI in medical imaging segment is estimated at USD 524.42 million in 2024, and is expected to expand at ~33.24% annually through 2030.

Those are big numbers, but what they really mean is this: every percentage point of efficiency you squeeze from imaging workflows can represent millions in saved costs, better throughput, and fewer compliance surprises.

That does not mean you simply “turn on AI.” You have to architect, govern, and integrate it in a way that maps to your enterprise’s scale, risk constraints, and growth agenda. This is why many leaders now prefer AI healthcare solutions built with enterprise-grade control instead of piecemeal point tools.

It is also why choosing custom healthcare software development becomes more than a preference. Your RIS must reflect your workflows, data flows, reporting priorities, and compliance path, not force your clinicians to adapt to someone else’s template.

So if you are wondering how much performance uplift is possible, how fast you can move into pilot, or how tightly you can tether it to regulatory guardrails, this guide is built for you.

What is an AI Radiology Information System (RIS)?

For any hospital or imaging enterprise, the Radiology Information System is the engine that keeps daily operations moving. It manages patient scheduling, tracks studies, organizes reports, and ties everything back to billing.

An AI Radiology Information System goes a step further by making these processes smarter and more proactive. Instead of simply recording information, it helps leaders see where delays, errors, or inefficiencies will surface and corrects them before they disrupt care.

For executives, this means:

• Fewer backlogs by anticipating workflow bottlenecks before they escalate
• Clearer compliance oversight with real-time audit and reporting visibility
• Faster turnaround times that improve patient satisfaction & enterprise revenue
• Scalable operations that connect RIS with PACS and EHR across multiple sites using enterprise AI solutions

In essence, an AI-powered RIS is not just an upgrade to legacy systems. It is a strategic asset that helps healthcare enterprises align clinical performance with business priorities, ensuring radiology departments contribute directly to growth, compliance, and patient trust.

How does an AI-powered Radiology Information System work?

An AI Radiology Information System functions as the central coordinator of imaging operations. Instead of simply storing records or tracking studies, it applies intelligence to routine tasks, guiding both clinical teams and administrators toward faster, more accurate, and more predictable outcomes.

1. Managing patient flow and scheduling

The system automates appointment handling, predicts no-shows, and allocates resources more effectively. This prevents bottlenecks and helps departments operate at full capacity.

2. Supporting diagnostic decisions

By analyzing imaging data and past reports, the system can flag anomalies, prioritize urgent cases, and reduce errors in documentation. Radiologists still make the final calls, but they do so with stronger support.

3. Providing enterprise visibility

Leaders gain consolidated dashboards that tie RIS data with PACS and EHR systems. This makes it easier to monitor performance, address compliance needs, and plan for multi-site scalability.

In practice, this means less time lost to manual errors and more consistency across the enterprise. With the right AI automation services, these capabilities can be scaled without overloading existing teams. Next, let’s look at why healthcare enterprises are choosing to invest in AI-driven RIS development in the first place.

Why Invest in the Development of AI Radiology Information System?

For healthcare enterprises, every significant investment must prove its worth in terms of clinical outcomes, operational efficiency, and financial performance. An AI Radiology Information System delivers on all three fronts, which is why it is fast becoming a board-level priority.

1. Stronger Financial Performance

Automating billing and reducing reporting errors directly decreases revenue leakage and claim denials. With cleaner data flowing through the system, financial leaders gain more predictable revenue and improved visibility during quarterly audits.

2. Faster Diagnostic Turnaround

AI-supported case prioritization pushes critical studies to the top of the queue while ensuring routine cases are handled efficiently. This results in shorter turnaround times, faster clinical decisions, and higher patient satisfaction across the enterprise.

3. Compliance Confidence

Built-in audit trails, secure handling of PHI, and automated monitoring allow leadership teams to maintain compliance readiness without pulling staff away from their core responsibilities. This reduces regulatory risk while strengthening governance practices.

4. Scalable Operations

AI-enabled RIS platforms adapt easily to multi-site rollouts and telemedicine programs. Partnering with a custom software development company ensures the system is tailored to grow alongside enterprise priorities rather than forcing future replacements.

With the benefits established, the next logical step is understanding the benefits of AI Radiology Information Systems available for enterprises.

Also Read: 50+ Questions to Ask Before AI Adoption in Healthcare

Key Benefits of AI Radiology Information System

For healthcare enterprises, the real question is not whether AI can improve radiology, but how much measurable value it brings across departments. An AI Radiology Information System directly impacts outcomes leaders care about most: patient experience, financial stability, compliance strength, and enterprise scalability.

1. Higher Patient Throughput

AI streamlines scheduling, reduces idle scanner time, and prioritizes urgent cases. Enterprises see shorter wait times for patients and more predictable daily volumes for staff.

2. Improved Diagnostic Quality

By surfacing anomalies and checking reports against prior studies, AI enhances accuracy without slowing radiologists down. This builds confidence with clinicians and minimizes the risk of costly re-reads.

3. Stronger Compliance Posture

Automated audit trails and PHI safeguards mean fewer surprises during inspections. Leadership gains visibility into where risks may surface before regulators point them out.

4. Enterprise-wide Efficiency

AI-powered RIS platforms integrate across multiple sites and service lines, eliminating silos. Many leaders rely on AI consulting services to align these systems with long-term enterprise growth strategies.

5. Optimized resource allocation

The system distributes workloads based on real-time demand and staff availability. This reduces burnout risk, controls overtime costs, and ensures high-value resources are used effectively.

6. Financial predictability

By tightening the billing cycle and reducing denied claims, enterprises stabilize revenue streams. CFOs gain greater confidence in forecasting and strategic planning, thus boosting overall financial operations.

Taken together, these benefits move radiology from a back-office function into a strategic driver of enterprise performance. With the value case established, the next step is to outline the must-have features that make an AI-enabled RIS truly effective.

Must-Have Features for AI Radiology Information System

Healthcare leaders evaluating an AI Radiology Information System often get distracted by futuristic add-ons, but what matters first are the foundational capabilities. These are the features that directly influence how well your enterprise manages patient flow, compliance, and revenue:

These must-haves can be achieved by working with teams who know how to integrate AI into an app while aligning with healthcare compliance standards. Enterprises that hire AI developers experienced in healthcare find it easier to embed these essentials in a way that scales with their growth trajectory.

Once the foundation is secure, the conversation shifts from “must-have” to “differentiator.” Next, we’ll explore the advanced features that can transform an AI-powered RIS from a reliable tool into a competitive edge.

Advanced Features to Build into Intelligent RIS Software

Advanced features in an AI Radiology Information System move beyond efficiency and create new opportunities for enterprises. They bring predictive insights, intelligent automation, and smarter patient engagement that help radiology departments become growth engines rather than cost centers.

1. Predictive workload balancing

Historical and real-time data are used to forecast imaging demand across departments. This ensures radiologists are never overloaded while equipment is utilized at maximum efficiency. Leaders gain predictability in managing high-volume workloads.

2. Natural language report generation

Structured and unstructured imaging data can be turned into standardized reports with minimal manual input. This reduces inconsistencies between radiologists and improves clarity for referring physicians. Administrative errors are also reduced.

3. Conversational AI for patient interaction

AI assistants handle routine tasks such as answering questions, booking appointments, and sending follow-up reminders. This reduces the administrative burden while improving patient satisfaction. Partnering with a AI chatbot development company makes adoption smoother.

4. Advanced compliance automation

Instead of only recording audit logs, the system monitors access patterns and data use in real time. It flags unusual activity that could signal compliance risks. Executives gain confidence in readiness for inspections.

5. Cross-department data orchestration

Radiology data can be integrated with oncology, cardiology, and surgical workflows to create a unified performance view. Leaders see how imaging impacts enterprise-level care delivery. Decisions are made with fuller context.

6. Generative AI for imaging insights

Generative AI offers support through comparative analysis, anomaly detection, and predictive simulations of disease progression. This helps clinicians make better-informed decisions without slowing workflows. Patient care outcomes benefit directly.

7. AI-powered financial optimization

The RIS predicts claim denial likelihood before submission and suggests corrections automatically. This strengthens revenue cycle management and improves cash flow reliability. CFOs gain clearer forecasting capabilities.

These advanced features are what separate a reliable RIS from an enterprise growth platform. They extend value beyond radiology and create measurable business impact across compliance, finance, and patient satisfaction. Next, we will map out the step-by-step process to create AI Radiology Information Management Software that delivers on these capabilities.

Step-by-Step Process to Create AI Radiology Information Management Software

Building an AI Radiology Information System for healthcare enterprises is not just about code. It is about aligning patient care goals, compliance needs, and enterprise efficiency into one scalable solution. Here’s a strategic breakdown of how leaders can navigate this journey.

1. Market Research and Planning

Every strong RIS begins with clarity. Enterprises must identify whether current bottlenecks stem from appointment scheduling, compliance blind spots, or lengthy reporting cycles. This stage sets priorities so the RIS delivers measurable business value instead of being a costly experiment.

• Conduct workshops with radiology heads, compliance officers, and financial leaders.
• Prioritize workflows to automate, such as scheduling, report generation, or billing.
• Validate early against HIPAA and other data protection standards.
• Define KPIs like reduced turnaround times or improved revenue cycle speed.

2. UI/UX Design

For physicians and technicians, user experience determines adoption. A clean interface ensures RIS tools become second nature in busy diagnostic settings. Partnering with a UI/UX design company helps enterprises build workflows that resonate with clinical and business users alike.

• Create prototypes simulating real diagnostic and administrative flows.
• Test designs with radiologists, technologists, and compliance staff.
• Add accessibility options like multi-language support and voice assistance.
• Maintain consistent design systems for cross-device usability.

Also read: Top UI/UX design companies in USA

3. Core Engineering and MVP Development

The smartest path is to start small with MVP services rather than over-engineer from day one. A lean rollout can include automated scheduling, PACS/EHR integration, and AI-assisted reporting. This ensures value is delivered early while giving leadership a chance to validate ROI.

• Build scheduling and reporting modules as the foundation.
• Enable AI-assisted data entry and report validation.
• Design the backend for scalability and cloud-readiness.
• Keep the roadmap flexible to integrate advanced AI later.

Also read: Custom MVP Software Development

4. AI and Data Integration

The “intelligence” in RIS comes from how data is processed. Clean pipelines and carefully trained models are essential for avoiding noise and bias. From predicting imaging demand to surfacing anomalies, AI integration must serve the enterprise vision rather than overwhelm it.

• Train models on diverse imaging datasets to minimize diagnostic bias.
• Implement continuous learning based on physician feedback.
• Develop anomaly detection systems tailored to enterprise patient mix.
• Balance pre-trained models with custom AI model development efforts for radiology-specific needs.

5. Security, Compliance, and Testing

Healthcare enterprises cannot compromise on trust. RIS platforms handle highly sensitive imaging data, so rigorous security and compliance are non-negotiable. Executives must demand transparency at every stage of testing to safeguard patient trust and regulatory standing.

• Conduct HIPAA compliance audits during development.
• Perform stress tests for high imaging workloads.
• Simulate user journeys to identify hidden usability gaps.
• Maintain detailed logs and access controls for accountability.

Also Read: Software Testing Companies in USA

6. Deployment and Cloud Readiness

RIS systems must scale to enterprise growth, mergers, or sudden patient surges. Cloud deployment makes this possible, ensuring reliability even during peak loads. Modern CI/CD practices also allow faster delivery of new features without disrupting hospital workflows.

• Deploy RIS on scalable cloud infrastructure.
• Implement CI/CD pipelines for rapid rollouts.
• Monitor enterprise usage through admin dashboards.
• Create onboarding playbooks for clinicians and staff.

7. Post-Launch and Continuous Optimization

No RIS is ever “done.” Continuous optimization ensures AI models stay accurate and enterprise value compounds over time. Post-launch, the focus should shift toward advanced analytics, predictive features, and staff training to maximize adoption.

• Collect ongoing feedback from radiology teams and administrators.
• Retrain AI models with fresh imaging data.
• Introduce advanced features like predictive scheduling or financial forecasting.
• Track ROI using business-centric KPIs such as patient throughput and revenue cycle speed.

By following these steps, healthcare leaders can reduce risks, accelerate ROI, and ensure their investment becomes a strategic differentiator. With the process clear, the next step is to finalize a suitable tech stack.

Also Read: A Step-by-Step Guide for AI Medical Software Development

Recommended Tech Stack for AI Radiology Information System Development

Choosing the right tech stack is about more than software preferences. For healthcare enterprises, the foundation must be scalable, compliant, and ready to support AI workloads that directly impact patient care and operational efficiency. Below is a comprehensive breakdown of technologies tailored for AI Radiology Information System development.

A carefully designed tech stack is the backbone of any successful AI-powered RIS, ensuring compliance, scalability, and intelligence all align with enterprise needs. With the right technologies in place, the natural next step is to evaluate what it costs to build AI-powered RIS software, from MVP through full enterprise deployment.

Cost of Developing AI-Powered RIS Software: MVP, Mid-level, Enterprise

Building an AI Radiology Information System is a strategic investment, and cost depends heavily on scope, complexity, and compliance needs. On average, healthcare enterprises can expect a ballpark range of $50,000 to $300,000+, depending on whether they’re starting with a minimal viable product (MVP) or scaling to an enterprise-grade RIS that integrates advanced AI and multi-site interoperability.

Here’s a quick breakdown of costs for a better understanding:

While cost is a critical factor, healthcare executives often view RIS development as less of an expense and more of a long-term investment in efficiency, compliance, and patient care. Once budgets are clear, the focus shifts to monetization of AI-powered Radiology Information System to maximize returns for enterprises.

Also Read: Cost of Implementing AI in Healthcare

How to Monetize AI Radiology Information Management Software?

An AI Radiology Information System is not just a cost center. When designed strategically, it becomes a profit engine by unlocking new revenue streams, reducing operational waste, and creating value that goes beyond diagnostic speed. Here’s how healthcare enterprises can monetize their RIS effectively.

1. Streamline Reimbursements and Reduce Claim Denials

Automated coding and billing powered by AI reduce errors that lead to costly denials. Faster reimbursements directly improve cash flow and free financial teams from manual rework.

2. Offer Subscription-based Imaging Services

Hospitals and diagnostic chains can package AI-enabled imaging reports as premium services. These subscriptions provide predictable recurring revenue and differentiate providers in competitive markets.

3. Monetize Predictive Analytics for Enterprise Partners

Imaging data, when anonymized and aggregated, can be turned into insights for insurers or research partners. Working with experts in AI model development ensures these analytics remain accurate and ethical.

4. Expand into Tele-radiology Markets

AI-powered RIS platforms support remote diagnostics, allowing enterprises to serve geographies where radiologist availability is limited. This opens new revenue channels while maximizing staff utilization.

5. Boost Patient Engagement Through Value-added Tools

Offering AI-driven chatbots, personalized notifications, or follow-up reminders improves patient retention. Higher retention translates into more repeat visits and stronger lifetime value per patient.

Monetization opportunities prove that an AI-powered RIS can generate revenue as well as savings. But maximizing returns requires smart execution, which brings us to the next step: the best practices healthcare enterprises should follow when developing an AI Radiology Information System.

Best Practices for Developing an AI Radiology Information System

For healthcare enterprises, building an AI Radiology Information System is about ensuring the system works seamlessly across clinical, financial, and regulatory dimensions. Executives who anchor development to these best practices set their organizations up for stronger adoption, measurable ROI, and long-term scalability.

1. Embed Compliance into Radiology Workflows

HIPAA, FDA, and enterprise-level compliance must be designed directly into RIS processes such as reporting and image archiving. This avoids expensive retrofits and gives leaders confidence during audits.

2. Prioritize Radiologist and Clinician Adoption

A sophisticated RIS is useless without buy-in from those who use it daily. Early usability testing with radiologists ensures the system reduces reporting friction rather than adding complexity.

3. Adopt a Modular Architecture for Growth

Healthcare enterprises often expand into tele-radiology or multi-site networks. A modular RIS architecture supports new service lines without disruptive rebuilds, keeping operations efficient at scale.

4. Leverage AI to Improve Diagnostic Accuracy and Revenue

AI should directly enhance outcomes by reducing reporting errors, accelerating reads, or improving claim approvals. Many leaders explore business app development using AI to align innovations with enterprise goals.

5. Plan for Ongoing Model Retraining and Optimization

Imaging data evolves, and so must your RIS. Continuous retraining of AI models ensures diagnostic suggestions remain accurate and workflows reflect the realities of modern healthcare enterprises.

While these practices ensure that the system becomes a growth engine. Still, even well-planned RIS projects encounter hurdles. Next, we will examine the challenges of AI Radiology Information Software Development and how enterprises can overcome them.

Challenges in AI Radiology Information Software Development and How to Overcome Them?

Healthcare executives know that building an AI Radiology Information System for enterprises is not a straight road. It comes with hurdles that directly impact patient care, diagnostic trust, and enterprise efficiency. The table below outlines the most common RIS-specific challenges leaders face, along with practical strategies to address them.

RIS development challenges are the roadblocks that decide whether enterprises achieve ROI or stall mid-project. The good news is each can be managed with foresight. Having said that, now let’s look at the regulatory considerations that shape AI Radiology Information System development in the U.S. healthcare market.

Regulatory Considerations in AI Radiology Information System Development

For healthcare enterprises, building an AI Radiology Information System is a high-stakes compliance journey. Every scan, report, and patient record processed through RIS carries regulatory weight. Leaders who address compliance early protect their organizations from legal exposure, reputational risks, and costly delays. Here’s how:

1. HIPAA and Patient Data Protection

Radiology workflows involve vast amounts of Protected Health Information (PHI). Encrypting data, enforcing access controls, and maintaining detailed audit trails are non-negotiable. Enterprises must also prepare for periodic HIPAA audits to avoid fines and strengthen patient trust.

2. FDA Oversight of AI in Diagnostics

AI modules used for diagnostic support may fall under FDA’s Software as a Medical Device (SaMD) framework. Proactive alignment with FDA guidance shortens approval cycles. It also reassures executives, investors, and clinicians that the RIS meets safety and reliability benchmarks.

3. CMS and Reimbursement Compliance

Without alignment to CMS billing and coding standards, enterprises risk delayed or denied reimbursements. AI-powered RIS must automate accurate coding and reporting. This ensures financial sustainability while reducing administrative burdens on clinical staff.

4. Interoperability Mandates (ONC, FHIR, HL7)

Federal rules increasingly demand data sharing across EHRs, PACS, and RIS platforms. Designing systems around standards like HL7 and FHIR enables smooth integrations. It also prevents costly rebuilds later when compliance enforcement becomes stricter.

5. State-Level Variations in the U.S.

Each state may add its own layers of regulation on top of federal law. Enterprises expanding across state lines must adapt their RIS accordingly. Proactive compliance planning ensures smoother scaling without regulatory bottlenecks.

Many enterprises partner with a software development company in Florida or similar specialists who understand both enterprise software execution and the nuances of U.S. regulatory landscapes.

Regulatory foresight not only minimizes risks but also accelerates adoption by inspiring stakeholder confidence. With compliance addressed, enterprises can now look ahead at the 2026 trends shaping AI Radiology Information System development.

2026 Trends in AI Radiology Information System Development

The radiology landscape is changing rapidly, and by 2026, enterprises that invest in the right AI Radiology Information System capabilities will lead the curve. These are shifts already gaining traction that will reshape how imaging workflows are designed, scaled, and monetized.

1. AI-Powered Precision Diagnostics

Enterprises will increasingly demand RIS platforms that move beyond efficiency to actually improve diagnostic accuracy. AI models will be retrained on massive imaging datasets, making them capable of detecting anomalies earlier and with higher confidence.

2. Integration of Conversational AI in Imaging Workflows

Radiologists and clinicians will expect natural interactions with RIS through voice and chat interfaces. Early adoption of conversational tools, like those mentioned in our healthcare conversational AI guide, will streamline reporting and reduce administrative bottlenecks.

3. Cloud-Native Expansion Across Enterprises

Multi-site healthcare groups will lean heavily on cloud-first RIS deployments. This approach will allow tele-radiology, remote collaboration, and rapid scaling into new regions without the downtime of traditional system upgrades.

4. Value-Based Care and Predictive Analytics

RIS will increasingly embed predictive analytics that tie directly to enterprise financial models. Predicting imaging demand, optimizing staff schedules, and reducing claim denials will become revenue-critical features for large organizations.

5. Ethical and Explainable AI Standards

By 2026, enterprises won’t settle for generic AI solutions in radiology. Explainable AI will become the norm, ensuring clinicians understand system recommendations, which strengthens trust and mitigates liability risks.

By aligning RIS development with them today, leaders can future-proof their investments. The final step is choosing the right partner who can translate this vision into reality, and that’s where Biz4Group’s expertise comes in.

Why Choose Biz4Group to Develop an AI Radiology Information System for Healthcare Enterprises?

At Biz4Group, we don’t just talk about building enterprise-ready AI solutions - we deliver them. As an AI App development company, our portfolio proves our ability to merge AI innovation with healthcare usability at scale.

Semuto: Personalized Healthcare Recommendations at Scale

Semuto empowers users with curated health and fitness app recommendations tailored to individual needs. Its ability to adapt to user data and provide meaningful insights mirrors the personalization required in intelligent RIS platforms, where radiologists need AI to guide precise, data-backed decisions.

Truman: AI-Powered Health Companion With a Human Touch

Truman introduced an AI avatar delivering wellness advice and tracking health history. This project highlights how AI can gain user trust while handling sensitive information—an expertise directly relevant to enterprises building AI-powered RIS software for radiology teams.

Healthcare enterprises need a technology partner who knows how to build AI software that’s scalable, compliant, and embraced by users. Biz4Group has done it across healthcare, and we’re ready to help you lead in radiology.

With proven success in AI healthcare applications and deep enterprise expertise, Biz4Group is uniquely positioned to deliver RIS platforms that balance compliance, innovation, and adoption.

Conclusion: AI Radiology Information System Development for Healthcare Businesses

Developing an AI Radiology Information System is a strategic investment that shapes how enterprises deliver care, manage costs, and stay competitive in a demanding healthcare landscape. With radiology at the heart of clinical decision-making, an intelligent RIS becomes the difference between keeping pace and leading the industry.

At Biz4Group, we’ve proven time and again that we’re builders of intelligent healthcare ecosystems. As an AI app development company, we help enterprises not only integrate AI but also harness it to solve business-critical problems. Whether it’s streamlining imaging operations, improving diagnostic accuracy, or ensuring compliance from day one, we’re here to help you turn the RIS of the future into your competitive advantage today.

See It In Action - Request a personalized demo and discover what an AI-driven RIS can do for your enterprise.

FAQs About AI Radiology Information System Development

1. How long does it take to develop an AI Radiology Information System?

Most healthcare enterprises can expect a timeline of 6–12 months depending on complexity. An MVP with core features may be ready in as little as 3–4 months, while a full enterprise solution with advanced AI and integrations will require longer development cycles.

2. What kind of ROI can healthcare enterprises expect from an AI-powered RIS?

ROI comes in multiple forms - faster turnaround times for reports, fewer diagnostic errors, reduced claim denials, and increased patient throughput. Enterprises often see operational efficiency gains within the first year of adoption.

3. What is the typical cost to develop an AI Radiology Information System?

The cost generally ranges between $50,000 and $300,000+, depending on scope. Factors like compliance readiness, AI feature depth, cloud architecture, and integration with existing systems can significantly influence the budget.

4. How secure is patient data in an AI Radiology Information System?

Security is central to RIS development. Encryption, access controls, audit trails, and HIPAA compliance checks are embedded into the design to ensure that Protected Health Information (PHI) remains secure and auditable.

5. Can AI Radiology Information Systems integrate with existing hospital systems?

Yes, modern RIS platforms are built with interoperability in mind. They integrate with EHRs, PACS, and billing systems using standards like HL7 and FHIR, ensuring smooth data exchange without workflow disruption.

6. What are the biggest risks of not adopting an AI Radiology Information System?

Enterprises that delay adoption risk falling behind on efficiency, diagnostic accuracy, and compliance readiness. This can lead to longer turnaround times, higher costs, reduced patient satisfaction, and a competitive disadvantage in the healthcare market.

Meet Author