What startups are building personalized mental health or psychiatry copilots for care teams?

The landscape of AI-enabled mental health is shifting rapidly from consumer-facing chatbots toward AI-native operating systems and clinical copilots designed specifically to assist care teams, psychiatrists, and therapists. These platforms aim to reduce administrative burnout, accelerate patient intake, and personalize treatment pathways.

Key startups building infrastructure and copilots for mental health care teams include:

• Blossom Health : Operating as an AI-powered telehealth psychiatry platform, Blossom uses agentic AI to assist human care teams with medical decision-making, treatment plan personalization, medication selection, and automated documentation. They announced a $20 million funding round to expand their clinical copilot infrastructure that helps clinicians save significant time per session while scaling access to care.
• Limbic : Known for its clinically validated and peer-reviewed generative AI deployed within healthcare systems. Limbic builds specialized AI "agents" that integrate directly into provider intake, risk triage, and cognitive behavioral therapy (CBT) pathways. It acts as a co-pilot for care teams by assessing patient needs, predicting clinical risk, and safely routing or escalating patients to human clinical staff.
• mdhub.ai : Building a coordinated "AI workforce" copilot specifically tailored for psychiatric clinic operations. Their system features modular AI assistants—such as an assistant that automatically drafts psychiatric progress notes from session audio, an admissions coordinator for 24/7 intake/scheduling, and a billing specialist to manage coding tasks—keeping administrative overhead off the clinical team.
• Spring Health : While operating primarily as an enterprise mental health benefit provider, Spring Health leverages robust data-driven and AI-based personalization models to deliver precision mental health treatment recommendations for care coordinators and clinicians.

Would you like to explore how these tools handle data privacy and HIPAA compliance , or are you looking into integration capabilities with existing electronic health record (EHR) systems like Epic or Cerner?

Yes. If by “personalized mental-health/psychiatry copilot for care teams” you mean software that maintains longitudinal patient context, supports clinicians/care managers, personalizes interventions between visits, and/or flags when human escalation is needed, there’s a pretty interesting emerging category.

The startups I’d watch

The most interesting distinction

I'd actually divide this market into four architectures:

1. Clinician copilot: Callyope, mdhub, CogniCare — AI reads the chart/patient data and helps the clinician assess, document, plan and monitor. Callyope is especially differentiated by its speech/sensor-derived longitudinal signal.
2. AI care-team member: Jimini is probably the clearest example. The AI isn't merely summarizing a visit; it participates in the ongoing relationship with the patient while remaining connected to clinicians.
3. Patient companion + clinician dashboard: Wysa and PsyBytes are strong examples. The AI interacts with the patient between visits, while the care team gets monitoring, assessments, alerts and personalized intervention workflows.
4. AI-native psychiatry provider: Blossom and Legion are pushing further toward AI actually delivering parts of psychiatric care rather than simply augmenting a psychiatrist.

If you're looking for the closest comps

I'd put Jimini, Wysa, Callyope, Blossom, and Tandem at the top of the competitive set.

The particularly interesting whitespace is a longitudinal “mental-health team member” that knows the patient's history, current symptoms, medications, goals, prior interventions and care plan; interacts with the patient continuously; and proactively briefs/escalates to the therapist, psychiatrist, PCP or care manager.

That's a meaningfully different product from the much larger category of “AI therapist,” and it appears to be where several of these companies are converging.

If you're evaluating this as a startup/market map, I can also break these down by funding, founders, investors, traction, target buyer, clinical workflow, AI architecture, and likely competitive moat.

Startups building AI copilots, ambient scribes, and decision-support tools for mental health and psychiatric care teams generally fall into two categories: clinical documentation/workflow copilots and personalized psychiatry/treatment optimization platforms.

1. Specialized Behavioral Health & Psychiatry Documentation Copilots

These startups build ambient AI assistants that listen to or process psychiatric and therapeutic sessions to draft compliant clinical documentation (SOAP, DAP, BIRP, and Mental Status Exam notes), track longitudinal progress, and reduce administrative burdens for multidisciplinary care teams.

• • JotPsych : Purpose-built specifically for psychiatry and behavioral health practices. It handles complex psychiatric documentation, medication management encounters, and coding support while integrating with existing EHR systems without requiring full migration.
• • Mentalyc : Tailored for therapists, counselors, and psychiatric nurse practitioners (PMHNPs). It generates progress notes, extracts specific symptoms and goals, and assists with treatment planning and tracking longitudinal client progress.
• • mdhub : An AI co-pilot designed for behavioral health clinics that coordinates a connected AI workflow spanning clinical documentation, scheduling, intake, and billing.
• • PMHScribe : Designed explicitly for psychiatric mental health nurse practitioners (PMHNPs), psychiatrists, and therapists to generate structured psychiatric evaluations and standard therapy notes.
• • Upheal : A modern, AI-driven platform for mental health practitioners that analyzes video or audio sessions to handle note-taking, provide analytics on communication patterns, and assist with clinical insights.

2. Personalized Psychiatry & Treatment-Decision Copilots

These companies focus more on the clinical decision-support and precision-medicine side, attempting to provide data-driven personalization for medication selection and treatment mapping.

• • Taliaz : Creator of PREDICTix, this platform uses machine learning and clinical data to support psychiatrists in personalizing antidepressant selection and predicting patient response, reducing the trial-and-error cycle in psychopharmacology.
• • Academic-backed and health-system frameworks (such as Stanford's CREATE center) are actively developing human-assist LLM tools to help care teams simulate therapy scenarios, implement evidence-based PTSD treatments, and monitor between-session patient data safely.

If you're looking into a specific workflow, let me know:

• Are you looking for ambient documentation/scribes or treatment/medication decision-support?
• Does it need to integrate with a specific EHR (like Epic or Valant)?

I can help narrow down the ideal platform.

Yes. There’s a fairly interesting emerging category here, but I’d separate true personalized care-team copilots from the much larger market of AI scribes / intake automation.

The startups I’d watch most closely

The most directly comparable ones

If by “personalized mental health/psychiatry copilot” you mean:

An AI that knows the individual patient, works continuously between encounters, gives the clinician a longitudinal view, recommends/executes appropriate interventions, and escalates to the human care team then I would put Jimini, Wysa, Limbic, Blueprint and Blossom at the top of the list.

Jimini is particularly close to the “AI member of the care team” concept. Its Sage assistant is explicitly designed to receive instructions from the clinical team, adapt support to a particular patient, monitor progress, and escalate situations back to providers.

Wysa has perhaps the clearest existing care-team infrastructure: clinicians can assign patient-specific tools, monitor assessments, communicate with patients, receive crisis flags and adjust care pathways.

Blueprint is more clinician-centric. Its assistant maintains client context and can reason about diagnostic questions, treatment planning, interventions and session preparation rather than merely producing a transcript.

Limbic is further upstream: it is building a clinical AI layer spanning intake → assessment → triage → therapy, with patient-specific care configured by providers. Its Limbic Care product lets clinicians customize activities and treatment plans for individual patients.

Blossom is the one I'd watch particularly closely on the psychiatry side. It raised $20M in March 2026 and is explicitly positioning its product as an AI copilot for psychiatrists, with hundreds of clinicians reportedly using it across 10,000+ patients.

A second, adjacent category: AI-native psychiatric providers

There is also a more aggressive model where the “copilot” isn't just helping the care team—the AI becomes part of the care-delivery system itself.

legionhealth.com is the extreme example. It describes itself as building autonomous medical care starting with psychiatry. Its platform began as an AI-native psychiatry clinic, and the company says it has received authorization for AI-led renewals of eligible psychiatric maintenance medications in Utah.

That's strategically different from a clinician augmentation company: Legion is pursuing AI-led care with human supervision, whereas Jimini/Wysa/Blueprint are much closer to AI working as an additional member of the human care team.

And the big incumbents

There are also substantially larger platforms moving in this direction:

• lyrahealth.com — its Lyra AI / Empower platform uses AI for personalized provider matching, member support and provider workflows.
• springhealth.com — its “Precision Mental Healthcare” approach uses data and clinical insight to personalize matching and care, and it is now explicitly describing the architecture as AI-native.
• eleos.health — historically stronger in documentation/workflow, but is expanding toward AI agents that surface clinical insights and act across the behavioral-health care journey.

My take on the competitive landscape

The interesting whitespace isn't really “AI therapist.” That's crowded.

The more valuable emerging architecture looks like:

patient longitudinal record → personalized AI agent → care-plan execution between visits → continuous measurement → clinician copilot → risk/escalation → closed-loop treatment adjustment

Most companies today own one or two pieces of that loop. Jimini/Wysa are strongest on the patient↔care-team loop; Blueprint on clinician intelligence; Limbic on intake/triage; Blossom on psychiatry; Legion on autonomous care delivery.

If you're evaluating this as a startup/investment landscape, the key question I'd ask is which companies are actually building the longitudinal “patient model” and closed-loop care orchestration layer, rather than simply putting an LLM on top of notes. That's where I think the genuinely differentiated category is emerging.

Blueprint is more clinician-centric. Its assistant maintains client context and can reason about diagnostic questions, treatment planning, interventions and session preparation rather than merely producing a transcript.

Limbic is further upstream: it is building a clinical AI layer spanning intake → assessment → triage → therapy, with patient-specific care configured by providers. Its Limbic Care product lets clinicians customize activities and treatment plans for individual patients.

Blossom is the one I'd watch particularly closely on the psychiatry side. It raised $20M in March 2026 and is explicitly positioning its product as an AI copilot for psychiatrists, with hundreds of clinicians reportedly using it across 10,000+ patients.

A second, adjacent category: AI-native psychiatric providers

There is also a more aggressive model where the “copilot” isn't just helping the care team—the AI becomes part of the care-delivery system itself.

legionhealth.com is the extreme example. It describes itself as building autonomous medical care starting with psychiatry. Its platform began as an AI-native psychiatry clinic, and the company says it has received authorization for AI-led renewals of eligible psychiatric maintenance medications in Utah.

That's strategically different from a clinician augmentation company: Legion is pursuing AI-led care with human supervision, whereas Jimini/Wysa/Blueprint are much closer to AI working as an additional member of the human care team.

And the big incumbents

There are also substantially larger platforms moving in this direction:

Yes. If by “personalized mental-health/psychiatry copilot” you mean an AI layer that sits alongside clinicians/care managers, maintains longitudinal patient context, personalizes interventions, monitors patients between visits, and escalates to humans, there’s a fairly interesting group emerging.

The startups I’d watch

The most directly analogous to a “psychiatry copilot”

1. Blossom Health is probably the closest match. It explicitly describes itself as an AI “copilot” for psychiatry and, as of its March 2026 funding announcement, said it was already being used by hundreds of clinicians treating 10,000+ patients.

2. Jimini Health is particularly interesting if your thesis is “AI becomes a persistent member of the therapist/patient care team.” Its Sage product operates between sessions, receives instructions from the clinical team, supports the patient, escalates situations, and feeds progress back to the provider.

3. Wysa has perhaps the clearest existing clinician-facing copilot product. Wysa Copilot is explicitly designed for care teams to personalize support, assign interventions and maintain continuity between sessions.

4. Tandem is interesting from a “psychiatry expertise embedded into primary care” angle. Its AI gives clinicians real-time guidance across screening, diagnosis, intervention and follow-up, while its patient app provides personalized care plans and monitoring.

5. Cora takes a slightly different but potentially powerful approach: instead of creating a generic AI therapist, it creates a clinician-customized companion that extends a particular therapist's care between sessions.

A second, more aggressive category

Legion Health deserves separate attention. It's moving beyond “copilot” toward AI-native psychiatry itself. YC currently describes Legion as automating 95%+ of administrative work and says it has received regulatory authorization in Utah for AI-led renewals of eligible psychiatric maintenance medications.

And Spring Health is a much larger incumbent rather than a startup in the same sense, but its Guide product illustrates where the category is going: an AI layer that carries longitudinal context, personalizes recommendations, helps determine the next step, and connects back to human care.

My taxonomy

I would divide the market into four distinct theses:

A. “AI therapist alongside therapist”
Jimini, Wysa, Cora

B. “AI clinical copilot for psychiatrist/care manager”
Blossom, Tandem, Limbus

C. “AI operating system for behavioral-health teams”
Psynk, aari

D. “AI-native psychiatric provider”
Legion, with Spring Health moving in that general direction at much larger scale

The most compelling whitespace, in my view, is between A and B: an AI that has a persistent patient-specific clinical memory, is explicitly configured by the treating psychiatrist/therapist, continuously gathers structured + conversational data between encounters, proposes the next best intervention, and gives the care team a concise “what changed / what matters / what should I do?” view.

That's substantially more ambitious than an AI scribe and safer/more clinically useful than an autonomous chatbot. It also creates a potentially defensible data/context layer around the care team rather than the individual patient.

The landscape of AI in mental health has shifted decisively away from standalone consumer chatbots trying to "replace" therapists, moving instead toward clinician-side and care-team copilots . These platforms augment psychiatrists, therapists, and care coordinators by handling documentation, surfacing clinical insights, and supporting patients between sessions under strict human supervision.

Key startups and platforms building personalized mental health and psychiatry copilots for care teams include:

• Eleos Health — Builds an AI-driven behavioral health platform specifically designed to capture and analyze therapy and psychiatric sessions . It automates clinical documentation and surfaces objective insights and progress metrics for care teams, helping them track treatment adherence and symptom changes over time.
• Limbic — A clinically validated generative AI platform used by health systems and providers . It functions as an intake, triage, and clinical workspace assistant . Limbic Care integrates medical AI scribing (via partnerships like Tortus) and caseload management into a unified dashboard, and it operates under strict regulatory frameworks, including deployments in supervised clinical environments.
• mdhub — Operates as an AI co-pilot and operating system explicitly for behavioral health and psychiatry clinics . It deploys specialized AI agents to handle time-consuming operational and clinical tasks—such as drafting psychiatric notes, managing intake data, and handling billing—leaving clinical judgment strictly to the human care team.
• Wysa — Well-known for its conversational CBT chatbot, Wysa has increasingly targeted hybrid clinical models . Its enterprise and provider toolsets act as a copilot for care teams by offering automated patient tracking, asynchronous between-session support, and early escalation alerts to licensed clinicians.
• Taliaz — Utilizes machine learning and data analytics to support psychiatrists in making personalized treatment and medication decisions . By gathering patient history, background, and previous responses, the platform helps streamline information-gathering and follow-up tracking to make psychiatric care more precise.
• LunaJoy — Focuses on precision mental health (starting with women's life stages) by embedding AI agents into care navigation, triage, and clinical decision support . It provides a hybrid model where AI assists the care team in personalizing and scaling psychiatric medication management and psychotherapy.

Would you like to explore how these tools handle HIPAA compliance and EHR integration , or do you want to focus on a specific workflow like automated clinical note-taking vs. patient-facing triage?

Startups in the digital health sector are shifting away from standalone consumer chatbots toward clinical-grade AI copilots and operating systems designed to augment multidisciplinary mental health and psychiatry care teams. Rather than replacing human clinicians, these platforms focus on personalizing treatment plans, automating administrative burdens, and tracking patient progress between structured visits.

Key Startups Building Mental Health & Psychiatry Copilots

• Blossom Health : Operating as an AI-native virtual psychiatry practice, Blossom raised $20 million in Series A funding to scale its AI operating system and clinical copilot for psychiatric teams. Their agentic workflow tooling assists providers across nine states with precision prescribing, automated documentation, prior authorizations, and patient care coordination.
• Limbic : A prominent clinical AI provider operating in both the UK (spanning large parts of the NHS) and the U.S., Limbic builds a regulated cognitive architecture for behavioral health teams. Their platform features intake/triage agents that predict diagnostic routing, and Limbic Care , an assistant for caseload management that integrates ambient voice scribing (via a partnership with Tortus AI ) to support clinicians between structured therapy or medication sessions.
• LunaJoy : Focused on precision and life-stage-specific mental health (such as reproductive psychiatry, postpartum care, and perimenopause), LunaJoy embeds AI agents into care navigation, triage, and clinical decision support . Their hybrid model helps coordinated care teams blend psychotherapy, genetic testing, and psychiatric medication management.
• Rose Health : Sells an end-to-end platform to physician groups and primary care systems. It deploys a daily patient check-in app utilizing natural language processing to analyze semantic tone in journals, feeding a unified dashboard for care teams and psychiatrists to monitor clinical risk and adjust pathways.

Would you like to explore how these companies handle compliance and HIPAA/Class II medical regulations , or look into enterprise mental health platforms like Spring Health and Lyra Health that use internal AI for care routing?

Yes. If by “personalized mental-health/psychiatry copilot” you mean software that sits alongside a clinician or care team, maintains longitudinal patient context, personalizes interventions/monitoring, and helps the team act—not just a consumer therapy chatbot—the field is getting interesting.

The startups I’d watch

The most direct comps

1. Blossom Health is probably the closest match if you're specifically thinking “AI copilot for a psychiatrist.” It describes an AI-native psychiatry platform where agents interact with patients between appointments, monitor things like mood/sleep, surface warning signs and tee up information for the psychiatrist. It raised $20M in combined seed/Series A in March 2026 and reported hundreds of clinicians and 10,000+ patients.

2. Wysa Copilot is perhaps the clearest example of “personalized AI layer around an existing care team.” Teams can assign interventions, monitor symptom trends and engagement, communicate with patients, detect risk, and adapt care pathways. Importantly, Wysa explicitly frames it as designed for clinicians, built for patients.

3. Limbic is particularly interesting if you're imagining the copilot becoming an agentic care-pathway layer rather than merely a scribe. Its stack now spans intake, triage and therapy, with Limbic Care described as an AI extension of a patient's treatment plan that engages patients between sessions.

4. Eleos is more clinician-centric. Its newer Clinical Insights Agent acts as an in-workflow copilot, combining client information with expert-vetted guidelines to support decisions around areas such as risk, evidence-based practice, substance use and CCBHC care.

5. Blueprint sits between AI scribe, longitudinal clinical memory and decision support. It can use prior sessions and assessments to help clinicians prepare, surface client progress, suggest evidence-based assessments/homework/interventions, and generate documentation. It reports 70,000+ mental-health professionals and 7M+ completed assessments.

An interesting second tier

Callyope is worth watching because its thesis is different: rather than primarily generating text, it is building a continuous psychiatric monitoring layer from speech, smartphone sensors and clinical information, aimed at detecting changes in symptoms such as depression, anxiety, insomnia and psychosis.

Legion Health is the more aggressive end of the spectrum. It's building an AI-native psychiatry service rather than simply selling a copilot to psychiatrists. In 2026 it announced scoped regulatory authorization for AI-powered renewals of eligible psychiatric maintenance medications in Utah.

The market seems to be converging on a particular architecture

The interesting pattern isn't really “AI therapist.” It's:

patient ↔ continuous AI layer ↔ human care team

where the AI layer:

1. maintains a longitudinal model of the patient;
2. checks in between visits;
3. monitors symptoms, behavior and risk;
4. personalizes homework/interventions;
5. summarizes changes for clinicians;
6. recommends next actions;
7. escalates exceptions to humans; and
8. gradually becomes an agent that executes low-risk workflow.

That distinction matters. The strongest companies are increasingly putting the clinician in the control loop rather than trying to replace the clinician. A recent npj Digital Medicine perspective similarly distinguishes assistive, collaborative and semi-autonomous psychiatric agents, with personalization and continuous monitoring among the most promising applications.

If you're doing a competitive landscape / startup scan, I'd put Blossom, Wysa, Limbic, Eleos, Blueprint and Callyope in the core set, with Legion as the “how far could this go?” outlier.

The landscape of AI tools supporting mental health and psychiatry care teams has shifted heavily toward ambient clinical copilots and specialized documentation scribes . Rather than trying to replace human therapists or psychiatrists (which faces major regulatory and clinical restrictions), these platforms function as co-pilots that handle documentation, triage, and workflow coordination.

The key startups and tech platforms building personalized mental health and psychiatry copilots for care teams include:

• Eleos Health : Focuses heavily on behavioral health agencies and care teams. It goes beyond basic note-taking by providing session-level insights, documentation automation, and analytics at scale for care organizations.
• JotPsych : Purpose-built specifically for behavioral health professionals, covering the full scope from traditional therapy sessions to complex psychiatric evaluations and medication management with customized templates and coding support.
• Limbic AI : A clinically validated generative AI platform used by health plans and care providers. It features specialized agents for patient intake, predictive triage, and structured escalation pathways back to human clinical staff.
• Medwriter.ai : Designed directly for psychiatric practices. It manages specialized tasks like Mental Status Exams (MSE), risk assessments, treatment planning, and E/M complexity scoring to optimize both clinical notes and reimbursement workflows.
• Mentalyc : Tailored for mental health clinicians to organize session recordings or notes into structured progress records (such as DAP and BIRP formats) to alleviate charting fatigue.
• Twofold Health : Widely recognized as a leading mental health-specific AI scribe that balances flexible note capture with behavioral health and psychiatry workflows, accommodating both solo practitioners and multi-provider group practices.
• Upheal : An all-in-one virtual telehealth platform that blends video sessions with automated clinical notes, client management, and behavioral insights for virtual-first care teams.
• Wysa : Corporate and enterprise-focused, Wysa has expanded its conversational CBT-trained AI agent models to bridge physical and mental health care coordination (including collaborative care models) while keeping human psychiatrists and care specialists looped in.

If you're looking into a specific deployment, tell me:

• Are you evaluating these for a solo practice, a large health system, or a digital health enterprise?
• Do you need integration with a specific EHR (like Epic or Valant)?

I can help narrow down which copilot fits your exact operational framework.

Yes. If by “personalized mental-health/psychiatry copilot” you mean software that sits alongside a clinician/care team, maintains longitudinal patient context, helps with assessment/treatment/monitoring, and/or gives the patient between-visit support, there’s a meaningful — but still relatively small — startup landscape.

The closest matches

1. Blossom Health — probably the most direct comp

Blossom Health is the one I'd watch most closely.

It describes itself as an AI-native psychiatry platform, and in March 2026 announced a $20M seed/Series A round led by Headline. Fortune reported that it was already being used by hundreds of clinicians treating 10,000+ patients.

The important distinction is that Blossom isn't merely an ambient scribe. The thesis is essentially AI-native psychiatric care, with AI taking on substantial portions of the workflow while clinicians remain involved.

Competitive position: closest to “AI copilot for a psychiatrist.”

---

2. Wysa — perhaps the clearest care-team + personalized AI architecture

Wysa has explicitly launched Wysa Copilot, designed for clinicians and care teams.

Its model is particularly relevant to your wording: clinicians can personalize support, assign specific tools/interventions, and keep patients engaged between sessions, while the patient interacts with AI.

That creates a useful architecture:

Clinician → personalized care plan → AI companion → longitudinal patient data → clinician

rather than:

Patient → generic chatbot

That's a significant distinction.

---

3. Frankly — interesting “psychiatric memory” play

Frankly is another particularly close match.

Its premise is that appointments compress weeks of psychiatric experience into a few minutes. Frankly is prescribed by the clinician and checks in with the patient between appointments, capturing symptoms and experiences for the care team.

I'd characterize this as:

AI longitudinal patient memory for psychiatry

rather than an AI therapist.

That's an interesting wedge because the product becomes more valuable as it accumulates patient-specific history.

---

4. Tandem — “psychiatry capabilities for primary care”

Tandem is attacking a different bottleneck.

It gives PCPs, OB/GYNs and women's-health clinicians AI-guided protocols spanning screening → diagnosis → treatment → follow-up, along with personalized patient care plans and between-visit monitoring.

This is important because the market may be much bigger than “AI for psychiatrists.” A lot of mental-health care happens in primary care and other medical specialties.

---

5. Eleos — the care-team infrastructure incumbent

Eleos Health is somewhat different.

Its original wedge was AI-assisted documentation and session intelligence for behavioral-health clinicians. But it's now moving toward AI agents that act on clinical and operational signals throughout the care journey.

So I'd put Eleos in the “behavioral-health care operating system” category rather than pure personalized copilot.

Its advantage is that it already has deep penetration into actual behavioral-health organizations and workflows.

---

6. Legion Health — the “copilot → autonomous psychiatrist” trajectory

Legion Health started with an AI-native psychiatry clinic in which psychiatrists and AI agents work together. It now describes a progression from clinician-led → AI-led with supervision → increasingly autonomous care for narrowly defined psychiatric workflows.

This makes it strategically different from Wysa or Frankly:

Wysa: AI extends the clinician's reach
Frankly: AI extends the patient's longitudinal monitoring
Blossom: AI extends the psychiatric care workflow
Legion: AI progressively takes over portions of psychiatric care

That distinction matters a lot.

---

7. Cognera Health

Cognera Health is building a broader continuous-care layer across mental health, behavioral health and wellness.

Its pitch is explicitly “continuous care beyond the session”, bringing together assessments, engagement, coordination, analytics and AI-generated clinical visibility while keeping licensed professionals as decision-makers.

I'd watch this one if you're interested in the infrastructure layer underlying personalized care rather than just the conversational interface.

---

8. April Health

April Health is an interesting hybrid.

It provides collaborative behavioral healthcare through behavioral-health care managers, while its Wysa-powered therapeutic coach provides between-session support and gives clinicians additional visibility.

It's less of a pure software startup and more of a tech-enabled care delivery model, but conceptually it is very close to the care-team-copilot idea.

---

How I'd map the market

The interesting thing is that these companies are actually pursuing four different versions of the same thesis:

1. “Copilot for the psychiatrist”

• Blossom
• Legion
• Limbus

2. “AI companion controlled by the care team”

• Wysa
• Frankly
• April/Wysa

3. “Mental-health intelligence layer for general clinicians”

• Tandem
• Cognera

4. “AI operating system for behavioral-health organizations”

• Eleos
• Cognera

Limbus is especially early: it describes itself as clinical AI infrastructure for psychiatric assessment, with psychiatrist review before diagnosis is finalized; its software medical-device authorization is still pending as of June 2026.

The white space I see

The really interesting gap isn't “ChatGPT for therapy.” That space is getting crowded and has substantial safety/clinical-validation issues. Recent work and reporting increasingly emphasize that human oversight is crucial for high-risk psychiatric use.

The more defensible product seems to be:

A longitudinal, patient-specific intelligence layer for the entire mental-health care team.

Something that continuously synthesizes:

patient conversations + symptoms + assessments + medications + adherence + clinician notes + patient-reported outcomes + behavior → personalized clinical picture → recommended next actions → human approval

That could serve the psychiatrist, therapist, care manager, PCP, and patient simultaneously.

In that framing, Blossom, Wysa, Frankly, Tandem and Eleos are probably the five companies I'd study most closely, but for different pieces of the stack rather than as identical competitors.