Presented by MongoDB
There’s a significant gap in the availability of care for children and adults diagnosed with autism and related intellectual and developmental disabilities (IDD) and who require clinical support. That ranges from applied behavior analysis and speech therapists to special educators, job coaches and more. Clinician and caregiver burnout is common, and the backlog of clients who need care is growing and organizations are struggling to deliver the necessary resources.
To bridge that gap, CentralReach has developed an AI-boosted autism and IDD care EMR (electronic medical record) platform that enables practitioners to deliver care more effectively — and to serve learners more productively and with greater results.
“One in 36 children are diagnosed with autism, and of those 2+ million children, less than half are estimated to be served today predominately because there aren’t enough clinicians to deliver care,” says Chris Sullens, CEO of CentralReach. “At CentralReach, we refer to this as the Autism and IDD Care gap and it drives our mission to deliver the best technology possible to compress the amount of time a provider’s team, especially the therapists, spends on administrative tasks so they can devote more of their attention to expanding the number of individuals they can help while also improving outcomes for those individuals. At the end of the day, the clinical solutions we have built and are building will improve efficiency of clinical staff, empower them to transform the way care is delivered and expand the number of individuals who receive care — making it a win for families, a win for the field, and, most importantly, a win for those children and adults we collectively serve.”
With over four billion clinical data points in their database across payers, states and service types, the company turned to MongoDB and the MongoDB AI Applications Program (MAAP) to help them build powerful AI-assisted solutions on top of that database. CentralReach’s goal was to help clinicians take advantage of the constantly growing body of client and treatment data, to unlock insights and to enable precise, individualized care.
“It surfaces all that information and puts it right at your fingertips. You don’t have to hunt and peck, or worry if someone filled out the right field or not,” says David Stevens, head of AI at CentralReach. “In that classic parlance, the AI doesn’t just find the needle in the haystack, it burns the haystack down. That’s our first use case.”
Building on this foundation, CentralReach is embarking on its Care360 initiative, which aims to unify these datasets into a comprehensive, 360-degree profile of each learner.
The solution leverages MongoDB’s flexible data model to integrate information from a wide range of providers — including board-certified behavior analysts (BCBAs) and speech therapists — into a single, holistic view of the client’s profile. It’s a unified approach that offers a more complete understanding of each learner’s needs while driving operational efficiencies with streamlined data access, reduced redundancies and enhanced care coordination across disciplines.
Looking ahead, CentralReach envisions extending the capabilities of Care360 by exploring real-time predictive analytics, AI-driven decision support and automated intervention recommendations. Future innovations may include dynamic care planning that adapts in real time to client progress, advanced natural language processing to analyze clinician notes and behavioral patterns and deeper integration with wearables for continuous, data-driven care. By building on MongoDB’s foundation, CentralReach is setting the stage for a future where AI-driven insights optimize operations and fundamentally enhance the quality of care for every learner.
To advance unstructured document ingestion, parsing, extraction and indexing capabilities for MongoDB’s document model, CentralReach has been working with MAAP partner LlamaIndex.
“Jerry Liu and his team’s expertise enables our AI to process and understand vast amounts of clinical documentation, making data more accessible and actionable, ultimately becoming a patient-specific oracle of clinical knowledge through a system of agents,” Stevens says.
With this buildout in 2025, CentralReach will be able to offer substantial improvements in clinical delivery and support initiatives like value-based outcome measurement, improved clinical supervision and greater care delivery efficacy.
“A document database like MongoDB Atlas allows you to do all that in a very flexible way that makes it more intuitive for the LLM to use and for us to use to feed them the right information,” Stevens explains. “And if the MongoDB database gives us agility with our data modeling and how we move from point A to point B, the MAAP program team are the sherpas that are taking us along for that journey.”
An evolution in EMR technology
At its core, the CentralReach platform is an EMR that combines both traditional practice management capabilities, like medical office software, claims processing, tracking and scheduling tools and so on, with client data collection and analysis. Analytics help assess progress and provide feedback and suggestions to clinicians. Every CentralReach tool is tightly integrated into that EMR to leverage that data, content and organizational expertise, while LLMs and AI automate and enhance workflows to reduce administrative burdens.
Finding a way for clinical directors and clinicians to talk to the data was a priority. These queries can include everything from simple updates — like when the client’s last assessment was, or how the client’s preferences have changed in the past week — to questions like “how is my client progressing on this skill?”
MongoDB AI tools enabled the creation of an agentic RAG system, which allows users to both ask questions of these document sets and learning trees, and to modify them by chatting with the CentralReach AI, cari. It updates documentation in a batch edit and presents them back to the clinician. The human guides the process and AI-in-the-loop applies edits (while infrastructure is kept behind the scenes), so users can focus on the clinical use cases and not the technology.
Underneath lies a huge amount of unstructured (but thematically related) data, from narratives to results from the clinician’s own practice, to data from providers across the country. MongoDB Atlas frees the data, structures it and adds a semantic layer of meaning on top with vectorization, summarization and classification tools, so that any user can make natural language queries across all these disparate data sources. Instead of a lengthy form-driven workflow to find information about a client’s preferences, the LLM can leverage MongoDB Atlas Vector Search to dig through the latest documentation and find the answer to that question semantically. And this all happens while keeping the confidential client data private and secure in MongoDB Atlas.
To help implement AI-driven retrieval and reasoning on clinical data, CentralReach worked with MongoDB MAAP partner gravity9 to develop data enrichment pipelines and Q&A agentic workflows that enabled scalable content retrieval and intelligent querying for unstructured healthcare data.
The pioneering MAAP ecosystem
Part of what drew CentralReach to MongoDB was the MongoDB AI Applications Program, which includes a network of tech leaders — including AWS, Google Cloud and Microsoft Azure — and AI innovators like Anthropic, Langchain and LlamaIndex. MAAP offers customers an array of resources to put AI applications into production, and the MAAP Center of Excellence Team — a cross-functional group of AI experts at MongoDB — collaborates with its partners and customers to overcome technical challenges.
“We’re overjoyed every time we get on the phone with some of these partners,” Stevens says. “They’re famous, the people that we work with. They’re out there pioneering the state of the art in how we do agentic applications, how we extract structured data from unstructured data. Being able to talk to the founders of Anthropic or LlamaIndex to get feedback on where we’re headed on our road map — just access to that level of people and companies doing this work — it’s unbelievable, the amount of support we can get.”
Pairing the expertise of MAAP network’s resources on how to most effectively leverage MongoDB toolsets with the expertise of CentralReach’s team of dedicated subject matter experts, clinicians and practitioners — and empowering them to fully utilize CentralReach’s clinical data set — unlocks extraordinary innovation and accelerates time to market, Sullens says.
“And ultimately, it positively impacts the lives of those we collectively serve,” he adds. “That’s why we’re both honored and energized by the ability to fully leverage MongoDB, the MAAP network and our unparalleled set of assets to accelerate the timeline for delivering on our Care360 vision.”
To that end, CentralReach’s next goal is to develop and commercialize their Care360 initiative.
“The idea is that the whole care team — including the caregivers, the parents and other people providing help outside of the clinical setting — have access to all this rich information about what’s working, what’s not working, are we on target to meet our goals,” says Stevens, “so the client is getting the most value for the time and funding available.” Together, CentralReach and MongoDB are demonstrating how AI can be used to create a meaningful impact for good.
To learn more about how MAAP can help organizations of all sizes make an impact with AI, check out the MongoDB AI Applications Program. And if you’re interested in hearing directly from MAAP partners, please visit the MAAP Ecosystem page.
Sponsored articles are content produced by a company that is either paying for the post or has a business relationship with VentureBeat, and they’re always clearly marked. For more information, contact
