In the last 6 months at @XYAILabs we:
✅ Launched our XY agentic platform with our first AI agents in production
✅ Built a world-class team with proven success in healthcare AI
✅ Onboarded our first customers with ROI already in motion
✅Raised our $3M seed round
Our mission is clear: give healthcare teams back the gift of time – so they can focus on what matters most, caring for patients.
Grateful to our investors and partners for believing in the vision.
#XYAI #XYAILabs #AgenticAI #HealthcareAI #HealthTech #RCM
Recapping the Digital Health & AI Innovation Conference, where Signal Labs CEO @RajeevRonanki moderated a discussion with @sam_debrouwer of XYAI Labs and Prem Somasundaram of Blue Cross Blue Shield of Massachusetts.
The conversation explored the future of work, automation, and AI’s evolving role across healthcare, highlighting the need to rethink processes rather than simply digitize them.
Great perspectives on where digital health and healthcare AI are headed.
#DHAI2026 #DigitalHealth #HealthcareAI
Healthcare runs on handoffs.
Shift to shift. AI to clinician. Payer to provider.
The future of work will be decided not inside any one role, but in the half-second between them.
Proud to sponsor #DHAI2026. @RajeevRonanki moderates Monday with @sam_debrouwer (@xyailabs) and Prem Somasundaram (CIO, BCBSMA)
AI can automate more healthcare decisions than ever. Which ones should still require a human? See you at #DHAI2026 with Signal Labs CEO @RajeevRonanki
to moderate that debate with @sam_debrouwer CEO
@xyailabs and Prem Somasundaram, CIO BCBSMA.
AI can automate more healthcare decisions than ever. Which ones should still require a human?
We're proud to sponsor #DHAI2026 Monday in Boston. Digital Health & AI Innovation Summit:
Signal Labs CEO @RajeevRonanki moderates that debate with @sam_debrouwer, CEO @xyailabs and Prem Somasundaram, CIO BCBSMA.
There’s a term we don’t talk about enough: token maxing.
It’s a simple idea. The more an AI tool works for a team, the more tokens it uses. The more tokens it uses, the more someone pays for compute. Some AI-native teams track it closely because it can show how deeply AI is being adopted.
But for healthcare operations, token maxing is the wrong goal.
Our agents already handle a large share of repeatable eligibility and benefits work, but not because we try to maximize token usage. It’s because we design the workflow carefully, then let the agent run the repeatable parts with controls, validations, and escalation paths.
And I don’t bill my clients that way.
If you run a mid-market healthcare practice like a surgery center, per-token pricing doesn’t work. Your volume spikes when patients show up, when payers tighten rules, when schedules change, or when codes need review. A token meter can turn every spike into a higher invoice.
So I compile the workflow at build time. The team designs it once, with the right rules, data checks, prompts, integrations, and human review points. The agent runs from there.
No bad surprises, just predictable pricing within agreed operating volumes, whether you process hundreds of checks or many thousands.
It’s not job replacement. It’s task compilation.
Finance can budget without surprises. Ops can focus on throughput and exceptions, not on defending the AI bill.
Token maxing may be the right signal for some AI teams. Compiled AI workflows are the right model for ours.
https://t.co/uViIhGHBQr
We’re excited to share that we're partnering with Opus to bring our AI agents to:
➡️10,000 behavioral health specialists
➡️ Serving 4M patients
Real workflows, automated. Details:
https://t.co/aqKoPjcDIE
#AI#HealthcareAI#BehavioralHealth#AgenticAI#HealthTech
AI in healthcare is hitting real scale.
Through our partnership with @findanuberdoc , https://t.co/TnTANYHQVQ is bringing AI agents to 5,000+ physicians across all 50 states in the US.
Real workflows, real impact.
👉 https://t.co/FuIe7Qd4VX
👉 https://t.co/nbzxEKnvgZ
A big, high-growth opportunity aligned with major societal needs. The “human-centered” angle offers room for innovation and leadership, but competitive.
From robotics to AI agents via care and precision medicine with the humans in the loop If you're shaping the future of care, this is the conversation you don't want to miss. Real use cases with AI in production.
#HealthcareAI#AIagents#AgenticAI#RCM
👏 👏 👏 Thanks for sharing @temporalio.
Our 1st orchestration system couldn't scale well when we needed to because:
Agentic AI in healthcare ≠ calling an LLM in a loop
With Temporal + a YAML DSL for our agentic workflows. Deterministic, scalable, debuggable.
What if your LLM could write the workflow instead of your engineers?
That's what XY built: a YAML DSL on top of a single Temporal Workflow class, where a Planner Agent generates healthcare automations from plain English. Claims processing, lab results, patient records all reliable enough to trust with PHI.
Learn how here: https://t.co/GLGhYrGDEF
What if your LLM could write the workflow instead of your engineers?
That's what XY built: a YAML DSL on top of a single Temporal Workflow class, where a Planner Agent generates healthcare automations from plain English. Claims processing, lab results, patient records all reliable enough to trust with PHI.
Learn how here: https://t.co/GLGhYrGDEF
Our first orchestration system couldn't scale well when we needed to because:
Agentic AI in healthcare ≠ calling an LLM in a loop
We moved to @temporalio + a YAML DSL for our agentic workflows. Deterministic, scalable, debuggable
Learn how here: https://t.co/W3td5JygSe
AI agents that compiles themselves out of existence? Our new paper in @arxiv_org introduces Compiled AI where LLMs generate the code once, then workflows run deterministically forever. 57x fewer tokens, zero runtime inference, 96% task completion👇
https://t.co/xS6VyTawTe
Excited to see our @XYAILabs CTO Lamara De Brouwer speaking at #NVIDIAGTC with @togethercompute.
Topic:
Healthcare Intelligence with Serverless Fine-Tuning at Scale
Or how AI infrastructure for real healthcare workflows.
🗓 March 18
⏰ 1 PM PST
#AI#LLM#HealthcareAI#GTC2026
We’re attending #MGMA Financial Conference (Mar 1–3, Phoenix).
Healthcare can’t hire its way out of structural inefficiency anymore.
It’s time to redesign execution.
If you’re going, let’s connect.
Details:
https://t.co/DLiyLnn6Fz
#Healthcare#AI#HealthTech#MGMA
Excited to join the 2026 AHA Rural Health Care Conference with 1,000 health leaders advancing care for their communities.
Our AI agents to save time, accelerate revenue to focus on what matters most.
🔗 https://t.co/Ol0YrixWCV
#HealthcareAI#RCM#AIagents#AIOps@ahahospitals
Great case study on @XYAILabs x @togethercompute 💡
Fine-tuning LLMs for complex EOB parsing, faster iteration, lower infra overhead, and better outcomes in healthcare AI.
This is what modern ML platforms should enable.
🔗 https://t.co/eGRPIFdpYY
We kicked off #JPMHealthcareWeek with a room full of builders, operators, customers, and curious minds.
Real conversations on:
• why agentic AI is ready for healthcare ops
• why it’s already working
• what comes next
Grateful to everyone who joined and contributed 🙏
📸👇