ASSYST, a leading innovator in IT and professional services, has been awarded a spot on the GSA OASIS+ Total Small Business (SB) Government-Wide Acquisition Contract (GWAC). This achievement reinforces ASSYST’s reputation for delivering agile, mission-driven solutions tailored to federal agencies’ unique operational needs.

“Securing a position on GSA OASIS+ Total Small Business is a testament to ASSYST’s ability to deliver transformative outcomes for our federal customers,” said Shyama Mandal, CEO of ASSYST. “Our innovative technologies, coupled with a deep understanding of mission-critical challenges, empower federal agencies with efficient, scalable, and impactful solutions.”

As a Total Small Business awardee, ASSYST is positioned to deliver high-impact services tailored to evolving federal missions. ASSYST's expertise spans areas such as enhancing agency performance, optimizing systems and infrastructure, advancing innovation in biotechnology and emerging technologies, and strengthening national security through cutting-edge tools and strategies.

The OASIS+ Total Small Business GWAC enables federal agencies to access various professional services while promoting small business participation. With a legacy of mission-centric solutions for Cabinet Level agencies such as the Department of Defense (DoD), Army, Energy, State Department, Health and Human Services (HHS), ASSYST is well-equipped to address challenges in Citizen Services, Health Systems, Energy Infrastructure, Financial Management, and National Security.

Green Accelerator Innovations, including Collab AI, ComplySyncAI, and Hephaestus, and investments in best-in-class technologies through partnerships, drive ASSYST's success. These innovations showcase ASSYST’s commitment to advancing federal
missions through leading-edge tools and strategies. ASSYST’s adaptive, agile approach ensures federal partners can overcome emerging threats, streamline operations, and foster innovation. The company remains committed to delivering exceptional value through tailored solutions that drive mission success. 

Learn More Discover how ASSYST’s small business capabilities can support your mission at www.assyst.net/oasisplussb or contact us at TeamUp@assyst.net

DISCOVER

Welcome to ASSYST OnPoint xChange! 

As we welcome the year 2025 Andrew Rios Account Manager at ASSYST , is discussing the convergence of IT Investments across Federal, State, and Local Governments with Ali Kalamchi Manager - Business Solutions. 

.... Continue Reading - Click Here (LinkedIn)

With millions of healthcare claims filed daily, even a small percentage of undetected fraud can result in significant financial losses for agencies. Traditional fraud detection systems often fail to keep up with the scale and complexity of modern claims processing, leading to inefficiencies, delays, and missed opportunities to protect valuable resources. Compounding this challenge, adhering to FHIR® (Fast Healthcare Interoperability Resources) standards introduces technical complexities that require advanced, scalable solutions.

Overview

In a Health and Human Services environment, managing healthcare claims' growing volume and complexity requires a robust and innovative approach to fraud detection. ASSYST developed an AI anomaly detection system to automate claims analysis, assign anomaly scores to high-risk cases, and deliver actionable insights through real-time monitoring and advanced dashboards. This solution addresses fraud detection challenges and ensures compliance with FHIR® standards, providing a secure, interoperable framework for seamless claims processing.

Challenges

The Health and Human Services sector faces several critical challenges in maintaining oversight and operational efficiency. Managing millions of claims overwhelms traditional detection systems, leading to delays in identifying fraudulent patterns and consuming valuable resources. Conventional rule-based methods struggle with accuracy, often generating high false positive rates and failing to adapt to new fraud tactics.

Adding to this complexity, FHIR® standards—essential for data standardization and interoperability—demand seamless integration into analytical systems. Agencies must balance compliance with the need for scalable, efficient, and adaptive fraud detection tools, a task that cannot be effectively managed with legacy systems alone.

Solution

ASSYST implemented an AI anomaly detection system that transforms claims processing and fraud prevention. The solution begins with automated pipelines validating, cleaning, and standardizing claims data by FHIR® requirements. This ensures the data is ready for analysis by a Decision Tree Classifier, which evaluates historical patterns, provider details, and claim attributes to detect anomalies. Each claim is assigned an anomaly score, prioritizing high-risk cases for further investigation.

Real-time alerts are generated to notify investigation teams of potential fraud, providing detailed indicators such as affected fields, anomaly likelihood, and supporting evidence. These alerts allow teams to focus on critical cases, improving efficiency and reducing resource waste.

The solution also features interactive dashboards that visualize key metrics, including fraud trends, detection rates, and operational performance. These dashboards give decision-makers a comprehensive, real-time view of claims data, enabling faster, data-driven actions. Adaptive machine learning capabilities ensure the system continuously evolves to identify emerging fraud patterns and improve detection accuracy.

Outcomes

The AI-enabled anomaly detection system delivered measurable improvements, transforming claims oversight and fraud prevention:

• Fraud detection timelines were reduced by 40%, allowing faster resolution of irregular claims.
• False positives decreased by 30%, freeing investigative resources for high-priority cases.
• Compliance with FHIR® standards ensured secure, standardized data processing, enabling seamless interoperability across systems.
• Operational efficiency improved significantly, with the system effortlessly scaling to handle millions of claims while optimizing resource utilization.

These outcomes equipped investigation teams with actionable insights and empowered decision-makers with the tools to adopt a proactive approach to fraud prevention, fostering greater trust and reliability in claims processing.

Conclusion

ASSYST’s AI enabled anomaly detection solution provides Health and Human Services agencies with a powerful framework for modernizing fraud detection and claims oversight. By leveraging FHIR® standards for seamless data integration and advanced analytics for anomaly detection, the solution accelerates fraud identification, reduces false positives, and enhances operational efficiency. Scalable and adaptive, this solution equips agencies to stay ahead of emerging fraud challenges, protect valuable resources, and deliver reliable, user-centric services.

ASSYST AI/ML Services

The ASSYST AI Center of Excellence (AICoE) plays a crucial role in this project. It partners with customers and their internal business and technology teams to embed AI expertise, enhancing agility and growth as AI matures. The CoE provides AI resources, strategy, and best practices, prepares AI Ready Data, and embeds data scientists, machine learning experts, and product managers into project teams to work closely with domain experts. Clear communication channels ensure ongoing collaboration and knowledge sharing. The CoE evaluates AI use cases, prioritizing them based on impact, feasibility, and strategic alignment. These use cases include automating repetitive tasks, enhancing customer experiences, improving communication strategies by identifying user behavior patterns, and detecting fraud or anomalies in financial transactions. 

We leverage the ASSYST Green Accelerator Program and solutions like Collab AI and ComplySyncAI to deliver accelerators that support these use cases. ASSYST focuses on data collection, preprocessing, and ensuring data quality and privacy. Model development leverages algorithms such as decision trees, support vector machines, and neural networks with precise training, validation, and performance assessment. The deployment process integrates models into production environments, leveraging a technology stack provided by Microsoft Azure, AWS, or other SaaS AI Tools. Continuous monitoring and maintenance ensure sustained accuracy, scalability, and reliability.

By incorporating and committing to human-centered design principles, we deliver intuitive, user-friendly, responsible, and humane AI solutions that meet end-users' needs.

Let’s discuss your AI Use Case.

Overview

While supporting the U.S. Health and Human Services (HHS) Agency, ASSYST, modernized its testing processes to overcome the limitations of traditional tools. Subtle visual and textual anomalies—like misaligned buttons, broken icons, and grammar errors—were frequently overlooked, slowing testing cycles and burdening developers. ASSYST implemented an AI-enabled solution that combined machine learning, object detection, OCR, and NLP techniques to streamline defect identification, reduce manual testing time, and enhance developer productivity. The outcome was faster testing cycles and defect-free digital services, improving the customer experience.

Challenges

The HHS Agency faced critical challenges ensuring a seamless and reliable user experience. Traditional testing tools struggled to identify subtle yet impactful visual defects, such as broken icons, alignment issues, layout inconsistencies, and unnecessary white spaces. Additionally, textual anomalies, including typographical and grammatical errors, required manual reviews that were time-consuming and error-prone. The growing scale and complexity of digital services further exacerbated these inefficiencies, slowing testing and delivery cycles. Developers spent excessive time identifying and resolving defects, delaying releases and diminishing productivity. The agency needed a precise solution to automate defect detection while accelerating the testing process.

Solution

To address these challenges, ASSYST implemented an AI-enabled defect detection solution that integrated machine learning models, OCR, and NLP. The solution was designed to automate the identification of visual and textual anomalies, ensuring faster and more accurate testing cycles.

A custom dataset of annotated UI images was created to train the Faster R-CNN ResNet-50 FPN model, leveraging transfer learning with PyTorch. This object detection model effectively identified visual defects, including broken icons, alignment problems, and white spaces across complex interfaces. To address textual inconsistencies, Tesseract OCR was integrated to extract on-screen text, which was analyzed using NLP libraries like NLTK and spaCy. This allowed the system to detect typographical errors, grammar issues, and content anomalies with exceptional accuracy.

The solution utilized OpenCV for image processing and integrated seamlessly into a .NET backend for real-time defect detection. Bounding boxes visually highlighted defects on live screens, enabling developers to identify and resolve issues quickly. The solution’s continuous learning capabilities also ensured adaptability to evolving UI designs, making it robust and scalable for future needs.

Metrics and Outcomes

The AI-enabled solution delivered significant improvements, transforming testing efficiency and accuracy:

• Manual testing time was reduced by 40%, accelerating defect identification and resolution.
• Visual defect detection achieved 95% accuracy, ensuring issues like alignment problems and broken icons were identified and addressed.
• Textual defect detection achieved 99% accuracy, efficiently resolving typographical and grammatical inconsistencies.
• Defects reaching production were reduced by 50%, ensuring a polished and seamless user experience.

These outcomes enabled faster testing cycles, improved defect resolution processes, and higher confidence in delivering error-free digital services.

Realized Benefits

ASSYST’s AI-enabled solution clearly benefited the HHS Agency’s development teams and end users. Developers could focus on innovation and core development work by automating repetitive and manual defect detection tasks. Testing processes were accelerated, boosting productivity and ensuring faster time to market for digital services. The solution’s ability to detect subtle visual anomalies and textual inconsistencies ensured a clean, professional, and error-free product.

For developers, this translated into greater efficiency and reduced frustration, leading to developer delight and improved team success. For end users, the seamless digital experience reinforced trust and satisfaction with the agency’s services.

Conclusion

While supporting the HHS project, ASSYST successfully implemented an AI-enabled defect detection solution that addressed the limitations of traditional testing tools. The solution automated defect identification reduced manual effort and accelerated testing cycles by integrating machine learning, OCR, and NLP. The result was improved developer productivity, faster defect resolution, and enhanced customer experience, positioning the HHS Agency to deliver reliable, user-centric digital services confidently.

ASSYST AI/ML Services

The ASSYST AI/ML Center of Excellence (CoE) plays a crucial role in this project. It partners with customers and their internal business and technology teams to embed AI expertise, enhancing agility and growth as AI matures. The CoE provides AI resources, strategy, and best practices, prepares AI Ready Data, embeds data scientists, machine learning experts, and product managers into project teams to work closely with domain experts. Clear communication channels ensure ongoing collaboration and knowledge sharing. The CoE evaluates AI use cases, prioritizing them based on impact, feasibility, and strategic alignment. These use cases include automating repetitive tasks, enhancing customer experiences, improving communication strategies by identifying user behavior patterns, and detecting fraud or anomalies in financial transactions. 

We leverage the ASSYST Green Accelerator Program and solutions like Collab AI and ComplySyncAI to deliver accelerators that support these use cases. ASSYST focuses on data collection, preprocessing, and ensuring data quality and privacy. Model development leverages algorithms such as decision trees, support vector machines, and neural networks with precise training, validation, and performance assessment. The deployment process integrates models into production environments, leveraging a technology stack provided by Microsoft Azure, AWS, or other SaaS AI Tools. Continuous monitoring and maintenance ensure sustained accuracy, scalability, and reliability.

By incorporating and committing to human-centered design principles, we deliver intuitive, user-friendly, responsible, and humane AI solutions that meet end-users' needs.

Let’s discuss your AI Use Case.

READ NOW

ASSYST is pleased to announce that we are now a Cybersecurity Maturity Model Certification (CMMC) Registered Practitioner Organization (RPO). This demonstrates our commitment to enhancing cybersecurity maturity and supporting organizations in achieving rigorous federal compliance standards.

Led by Mr. Vijay Narasimhan, CTO and a certified CMMC Registered Practitioner, ASSYST provides tailored guidance to help clients navigate the complexities of CMMC compliance. Vijay’s expertise and strategic vision are central to our mission.

https://cyberab.org/Member/RPO-62569-Assyst-Inc 

As we align with evolving cybersecurity frameworks, ASSYST is uniquely positioned to deliver comprehensive solutions that ensure secure, compliant operations for federal, state, and local organizations, fostering a future of security and trust across critical systems and data.

ASSYST is excited to announce our participation as a teaming partner in a winning bid for the Federal Aviation Administration (FAA) Information Technology Innovative Procurement Strategic Sourcing (ITIPSS) contract. This IDIQ multiple-award contract, with a $2.4 billion shared ceiling over ten years, will enable the FAA to acquire a full range of IT capabilities, solutions, and emerging technologies, providing state-of-the-art IT-related service solutions.

The ITIPSS contract is key for acquiring non-National Airspace System (non-NAS) IT services. It supports various functions, including FAA facilities inspections, internal and external investigations, IT resource management, financial management, and security.

Contact:

Eugene Goldlust, Sr. Account Executive

egoldluts@assyst.net | LinkedIn

In this detailed Q&A, Vijay Narasimhan, CTO at ASSYST, and John Kimberl, Health Data Analyst, discuss how AI—combined with the power of the FHIR® standard—is transforming health data interoperability. They delve into specific, highly technical use cases where AI enhances data discoverability, transformation, validation, and reporting, with each use case underscoring a step closer to achieving seamless health data ecosystems.

The 21st Century Cures Act has driven significant change by promoting interoperability, enhancing patient access to health data, and curbing information blocking practices. By mandating the adoption of FHIR®, the Act accelerates seamless data exchange across systems, improving care coordination. Moreover, focusing on empowering patients, the Cures Act opens the door for AI technologies to deliver personalized insights, predictive analytics, and dynamic care plans that enhance patient engagement and outcomes. As AI continues to innovate, the Act's provisions support the shift toward real-time, accessible, and actionable data—transforming the healthcare ecosystem.

John: Vijay, let’s dive into how AI’s technical capabilities are transforming health data interoperability, especially with FHIR® as a foundational standard. What role do you see AI playing in connecting disparate health data systems?

Vijay: AI is critical in creating adaptive, fully automated interoperability frameworks where FHIR® plays a pivotal role. With advanced machine learning models and FHIR®’s flexible data structure, we’re building platforms that harmonize diverse data formats into a universally accessible model. This is especially vital in environments like hospitals, research institutions, and public health agencies, where data must flow seamlessly between EHR systems, external labs, and other healthcare services. Using ASSYST’s Hephaestus FHIR® aPaaS, for instance, we can deploy an AI-powered interoperability solution that manages schema differences, provides real-time updates, and ensures standardized data exchange, thus delivering a holistic view of patient records.

Use Case: Enhanced Data Discoverability through AI

John: Could you talk about how AI-driven data discoverability works in a FHIR®-based ecosystem?

Vijay: Absolutely. AI revolutionizes data discoverability by using natural language processing (NLP) and entity recognition models trained on healthcare-specific terminologies like ICD-10 and SNOMED CT, seamlessly integrated with FHIR®-based resources. This enables semantic search capabilities, allowing clinicians or researchers to query vast datasets in natural language without needing deep technical knowledge of coding or terminologies.

When retrieving FHIR® resources across data sources, AI relies on FHIR® APIs, standardizing how structured health data is queried and exchanged. These APIs facilitate interoperability across disparate systems, whether within a single Health Information Network (HIN), across Health Information Exchanges (HIEs), or even between unrelated organizations, provided data-sharing agreements and permissions are in place.

AI orchestrates the process through several mechanisms:

• Federated Search Across HINs or HIEs: AI performs federated searches by querying multiple repositories simultaneously. It uses API endpoints to gather relevant resources, such as Observation or DiagnosticReport, based on specific query parameters, regardless of the data's physical location.
• Mapping and Normalization: AI ensures retrieved data adheres to a common schema, reconciling differences in terminologies or formats between sources. For example, AI maps local codes to SNOMED CT, enabling consistent and actionable insights.
• Security and Data Sharing Agreements: AI respects data governance frameworks like TEFCA, ensuring compliance with patient consent and privacy regulations such as HIPAA. It retrieves data only within the bounds of established agreements and permissions.
• Dynamic Resource Linking: AI creates contextual connections between resources. For example, when searching for “recent bloodwork with abnormal readings,” AI retrieves relevant Observation resources and dynamically links them to related Condition or Procedure resources within the FHIR® schema, simplifying the clinician's workflow.
• Caching and Real-Time Updates: AI caches frequently accessed data while querying live sources for updates to optimize performance. This hybrid approach ensures clinicians always access the most current and comprehensive information.

Use Case: Seamless Data Transformation Across Standards

John: Data transformation must be challenging, especially when healthcare data comes from various sources. How does AI handle these transformations within an FHIR® framework?

Vijay: Data transformation in healthcare is a complex task, given the range of formats like HL7, DICOM, and CDA. AI allows for adaptive mapping from these legacy formats to FHIR® resources. We use Transformer models customized for structured and unstructured data that automatically recognize and map data fields to FHIR® resources, such as mapping patient demographics from HL7 messages to FHIR® Patient resources or imaging data from DICOM to FHIR® ImagingStudy resources.

Moreover, these models improve their transformation accuracy over time using reinforcement learning. The models are trained to identify patterns and adapt mappings based on incoming data, making it possible to refine transformation processes continuously. For example, if a hospital system updates its coding practices or EHR schema, AI can detect and adapt to these changes autonomously, minimizing the need for manual intervention.

Use Case: Intelligent Data Validation to Ensure Integrity and Compliance

John: How does AI support data validation, especially given the need for strict compliance in healthcare?

Vijay: AI-driven data validation is essential for ensuring that health data remains compliant and clinically accurate. Our platform uses rule-based engines combined with ML models to validate data at multiple levels. For example, in a FHIR®-based system, AI checks if Patient resources have necessary fields filled, such as identifiers, and flags anomalies like invalid age ranges or out-of-bound values in Observation resources.

Our anomaly detection algorithms also work to ensure data integrity. By applying clustering techniques, the system can flag outliers or unexpected deviations in health metrics—for instance, identifying sudden spikes in patient vitals or dosage errors in medication data. This real-time validation helps in regulatory compliance and significantly reduces the chances of data errors affecting patient care.

Use Case: Real-Time Data Aggregation and Reporting

John: Could you describe how AI enhances real-time data aggregation and reporting within FHIR® ecosystems?

Vijay: AI’s role in real-time reporting and aggregation is transformative. By leveraging graph-based machine learning, AI can analyze relationships between FHIR® resources—such as linking Patient and Condition resources to identify population health trends. This allows healthcare providers to quickly access a consolidated view of patient data, helping them identify trends, risks, or treatment outcomes across various patient cohorts.

In practice, AI-driven tools can continuously aggregate data streams, automatically updating dashboards with the latest health data. Our AI-powered reporting system within a FHIR® environment also enables predictive analytics, such as predicting patient readmission rates based on trends in Observation and Procedure resources. The reports can be visualized dynamically, letting users drill down into metrics by diagnosis, location, or demographic data, which is invaluable for hospital administrators and public health officials.

Use Case: Automating Compliance with FHIR® and Regulatory Standards

John: Compliance is a big concern in healthcare. How does AI help ensure that FHIR®-based systems meet regulatory requirements?

Vijay: AI streamlines compliance by automating validation checks, maintaining audit trails, and proactively monitoring adherence to regulatory frameworks like HIPAA. It continuously validates FHIR® resources to ensure they meet structural and logical standards while also enforcing data-sharing policies. For instance, AI ensures that only the "minimum necessary" data is shared, as mandated by HIPAA, and flags any violations in real-time.

In addition, AI-driven anomaly detection strengthens compliance efforts by identifying irregularities in FHIR® data formats or transactions. For example, it can flag missing or unexpected fields in Patient or Observation resources or detect unusual transaction patterns, such as an unexpected spike in data requests that may indicate unauthorized access. These capabilities enable early detection of compliance risks or data integrity issues.

AI also simplifies audits by generating detailed compliance reports through natural language processing (NLP), making it easier to meet regulatory requirements. By embedding these capabilities directly into the FHIR® framework, our platform minimizes compliance risks while reducing manual oversight, ensuring data integrity and regulatory adherence across all transactions.

Use Case: Predictive Health Analytics with FHIR® Data

John: How does AI leverage FHIR® data to support predictive health analytics?

Vijay: Predictive analytics with FHIR® data relies heavily on AI to analyze historical patient data and model future health outcomes. Using time-series analysis and regression models on structured FHIR® resources, AI can predict outcomes like the likelihood of disease progression or potential readmission risks. For instance, AI models can process Condition, MedicationRequest, and Observation resources to identify patients at risk of chronic conditions, allowing for early intervention.

We’re also implementing reinforcement learning algorithms that dynamically adjust predictive models based on new data inputs, improving prediction accuracy over time. This helps healthcare providers allocate resources more effectively and implement preventive care measures based on AI-driven insights, ultimately enhancing patient outcomes.

John: It’s clear AI, combined with FHIR®, enables highly technical, transformative use cases across the healthcare spectrum. Thanks for these insights, Vijay.

Vijay: It's my pleasure, John. AI gives us the tools to enhance interoperability and optimize healthcare delivery through robust, reliable data. The integration of AI with FHIR® marks a new era in healthcare data, where seamless, compliant, and actionable insights are not just possible but are becoming the new standard.

ASSYST Health IT 

ASSYST brings over three decades of expertise in delivering innovative healthcare IT solutions to federal, state, and local government agencies. As a proud HL7 Gold Member, ASSYST is deeply committed to advancing health data interoperability, leveraging industry-leading standards like FHIR® to transform healthcare delivery.

Our extensive experience spans key agencies such as the ONC/ASTP, Centers for Medicare & Medicaid Services (CMS), the Centers for Disease Control and Prevention (CDC), the Food and Drug Administration (FDA), the Health Resources and Services Administration (HRSA), and the Program Support Center (PSC). Additionally, we have supported State and Local Health and Human Services (HHS) agencies in improving healthcare systems, ensuring compliance, and driving digital transformation.

ASSYST’s Green Accelerator powered Hephaestus aPaaS (https://www.assyst.net/hephaestus/) is a powerful platform accelerating health data integration and innovation. By embedding FHIR® standards, Hephaestus supports healthcare organizations in streamlining workflows and improving data discoverability, validation, and reporting while enhancing compliance with regulatory requirements. With a clear focus on interoperability, security, and operational efficiency, ASSYST’s healthcare practice empowers clients to meet the evolving demands of the healthcare ecosystem and achieve better health outcomes through technology-driven solutions.

ASSYST, a leading IT solutions provider, is pleased to announce its selection as a Prime Contract holder under Pool 2 of the U.S. Department of Agriculture (USDA) STRATUS Basic Ordering Agreement (BOA). The STRATUS BOA, with a performance period extending through 2034, offers USDA a modern platform for accelerating cloud adoption, modernizing IT operations, and advancing mission delivery through innovative cloud integration, application modernization, and cloud security services.

Read More...

ASSYST, a leader in advanced technology solutions, has been named a Northern Virginia Technology Council (NVTC) Tech 100 honoree. This prestigious recognition highlights the top companies and individuals making transformative contributions to the region’s vibrant technology ecosystem. The NVTC Tech100 celebration event will be held on December 10, 2024, from 6 to 8 PM at the Hilton McLean Tysons Corner, Virginia. 

Read More...