AI Cred profile
Costa Cassis
Founder & CTO Future Era AI
Phoenix, AZ
@ccassis
AI fluency score
9.5Expert
Assessed Jul 4, 2026 · 2026 rubric
The five skills
Strongest: Verification & Judgment
About
Costa builds AI systems that have to survive contact with a regulator, not just a demo. Over more than 20 years he has turned enterprise data into decisions for healthcare payers, providers, the public sector, and global retail, and he has done it as both the executive who sets the strategy and the engineer who ships the code. He builds the teams, platforms, and governance that move AI from a board memo into work people actually use, and when the moment calls for it, he architects and ships the platform himself. He currently leads enterprise AI at Medecision as its first dedicated AI leader, building the function from the ground up across both sides of the business: the population-health and care-management SaaS platform, and the AI advisory practice (formerly Excell Healthcare Advisors, acquired by Medecision in September 2025). He wrote the company's AI governance posture against regulation that already exists rather than theory: human-in-the-loop design, model documentation, and audit-response plans built against Alabama SB 63, California SB 1120, and the UnitedHealthcare nH Predict and Cigna PXDX cases. He defined the AI service offerings the company takes to market (readiness assessment, governance stand-up, strategy and roadmap, and workflow redesign across utilization management, prior auth, care management, and member services), and he drove adoption past engineering: expanding Claude access, turning on Copilot where it was licensed but switched off, and forcing a written AI use policy after the rules were being decided ad hoc in chat threads. He is also the founder of Future Era AI, where he built Lumina, a multi-tenant, AI-native analytics platform that lets a non-technical person ask a question in plain language and get back governed SQL, a chart, and a written narrative, without copying a byte of the customer's data. He architected the federated query engine himself on Trino and SQLAlchemy, answering questions across 17+ source types (Snowflake, BigQuery, Redshift, Databricks, Postgres, and more) by querying each source on demand. The natural-language-to-SQL pipeline runs on a Qdrant semantic layer with Claude and OpenAI behind a provider factory, and a query-time governance engine handles column masking, row-level filters, SQL-parsed lineage, immutable audit logging, and per-tenant credential encryption. Production runs on Azure Container Apps under a single BAA through Microsoft AI Foundry, which makes the platform usable against HIPAA-bound data. He also built Statute Watch, an agent that tracks state AI legislation affecting healthcare payers end to end: pulling the daily legislative digest, updating a curated bill-tracking workbook while preserving human-curated fields, and flagging new bills for review, unattended. Earlier, he was Managing Director of Data Intelligence at NTT DATA, where he directed enterprise data transformation for payer, provider, and public-sector clients, led a 150-plus person global delivery team, and oversaw $15M+ in realized value by integrating AI with revenue-cycle automation, public-sector grants, and care-management innovation. He launched a cloud-first data stack on Snowflake, dbt, and Airflow, decommissioned 25+ legacy systems and mainframes, and led predictive modeling that cut claims-review cycles by 65%. Before that, as Program Director at Anthem, he led a $5.5M enterprise API program to deliver ONC/CMS FHIR-compliant endpoints, and as Director of Solution Architecture at Unite Us, he cut the implementation time for a national social-determinants-of-health platform from six months to under eight weeks while integrating with Cerner, Epic, and Allscripts across 14-plus states. His earliest executive work was just as hands-on. As Enterprise Architect at Blue Cross Blue Shield of Arizona, he migrated legacy claims and eligibility systems to Azure PaaS, cutting infrastructure cost by $2M and call-center volume by 48%. As Program Director at River Valley Behavioral Health, he took over a three-year stalled EHR rollout and delivered five go-lives in 18 months. As IT Manager at Luxottica, a $66B global retail brand, he delivered a $400M application and infrastructure migration in 14 months and cut projected data-center cost by $130M over five years, while running an offshore support center in China for 10,000 stores. He holds a Master of Science from Northern Kentucky University and a Bachelor of Arts from Thomas More College. He is known for cutting the forty-box architecture diagram down to the three things that matter, and for designing the human-in-the-loop and audit posture that keeps an AI program defensible when a regulator asks to see the math. If he cannot personally defend the output, he does not delegate the action.
In short
Written by AI from Costa's assessment and profile, Jul 4, 2026.
Experience
AI Market VP at Medecision (Dec 2025 – Present)
Building Medecision's enterprise AI function from the ground up as the company's first dedicated AI leader, across the population-health/care-management SaaS platform and the AI advisory practice (formerly Excell Healthcare Advisors, acquired September 2025).
- Built the company's AI governance posture against existing regulation: human-in-the-loop design, model documentation, and audit-response plans written to Alabama SB 63, California SB 1120, and the UnitedHealthcare nH Predict and Cigna PXDX cases.
- Defined the AI service offerings taken to market: readiness assessment, governance stand-up, strategy and roadmap, change management, and workflow redesign across utilization management, prior auth, care management, and member services.
- Drove enterprise AI adoption beyond engineering: expanded Claude access, enabled Copilot inside Office, and forced a written AI use policy after ad hoc decisions in chat threads.
- Ran role-based AI enablement for non-technical teams across HR, project delivery, and operations.
- Authored an internal field manual for senior consultants on AI strategy, governance, and use-case identification.
Founder & CTO at Future Era AI (Jan 2025 – Present)
Founded Future Era AI and built Lumina, a multi-tenant, AI-native analytics platform that lets a non-technical user ask a question in plain language and get back governed SQL, a chart, and a written narrative, without copying a byte of the customer's data. Concurrent with the Medecision role.
- Architected a federated query engine on Trino and SQLAlchemy that answers natural-language questions across 17+ source types, including Snowflake, BigQuery, Redshift, Synapse, Databricks, Fabric, SQL Server, Postgres, Oracle, and MongoDB, querying each source on demand instead of copying tenant data into a warehouse.
- Built the natural-language-to-SQL pipeline on a Qdrant semantic layer of schema, table, column, and business-glossary embeddings, with Claude and OpenAI behind a provider factory.
- Designed a query-time governance engine (column masking, row-level filters, SQL-parsed data lineage, immutable audit logging, per-tenant credential encryption, and contract-expiration checks) to keep analytics defensible in regulated settings.
- Stood up production on Azure Container Apps, routing every LLM call through Microsoft AI Foundry under managed identity and a single BAA covering chat and embeddings.
- Shipped real-time streaming execution over Redis pub/sub and WebSockets plus multi-user collaboration, backed by a Celery worker fleet handling metadata crawling, embedding, scheduled reporting, and schema-drift detection.
VP, Data & Analytics at Excell Healthcare Advisors (May 2024 – Jan 2025)
Led AI-driven modernization of enterprise healthcare analytics, positioning data as a monetizable product aligned to clinical and operational goals. Practice acquired by Medecision in September 2025.
- Designed a secure, cloud-native architecture on Azure, dbt, and Databricks that unified structured and unstructured health data from ingestion through model operations, with HIPAA-compliant audit layers.
- Built an AI-enhanced reporting suite using prompt-based insight generation and real-time dashboards, replacing static workflows and increasing report usage.
- Established a roadmap for a secure, HIPAA- and SOC2-compliant AI data-sharing model, laying the foundation for token-based data provisioning to external partners.
Managing Director, Data Intelligence at NTT DATA (Oct 2021 – Apr 2024)
- Directed enterprise data transformation for payer, provider, and public-sector clients, leading a 150+ person global delivery team and overseeing $15M+ in value creation.
- Delivered $15M+ in realized value by integrating AI with revenue-cycle automation, public-sector grants, and care-management innovation.
- Launched a cloud-first data stack on Snowflake, dbt, and Airflow and decommissioned 25+ legacy systems and mainframes.
- Led predictive modeling for claims and care management, cutting review cycles by 65% and enabling proactive clinical intervention.
- Defined an enterprise GenAI adoption roadmap with security, ethics, and infrastructure policy, and built secure deployment environments with centralized MLOps.
Program Director at Anthem Inc. (Feb 2021 – Oct 2021)
- Led compliance and interoperability across a $5.5M enterprise API program, delivering ONC/CMS FHIR-compliant Patient Access, Formulary, and Provider Directory endpoints on an accelerated timeline.
- Managed a 15-person cross-functional team across architecture, data, QA, and infrastructure, delivering fully auditable FHIR APIs under budget and on schedule.
- Reduced provider onboarding time by 40% by rewriting authentication and OAuth flows, and automated real-time CMS audit-compliance reporting with dbt and Python.
Director of Solution Architecture at Unite Us (Dec 2019 – Jan 2021)
- Led architecture and delivery of a national social-determinants-of-health referral platform integrated with EMRs and community organizations, scaling across 14+ states.
- Cut implementation time from 6 months to under 8 weeks by modularizing onboarding, ETL, and schema-validation layers.
- Integrated with Cerner, Epic, and Allscripts using SMART on FHIR, SCIM, and SAML across multi-tenant deployments.
Enterprise Architect at Blue Cross Blue Shield of Arizona (Oct 2017 – Dec 2019)
- Led enterprise modernization across claims, member services, and provider networks, migrating legacy on-prem systems to Azure PaaS with real-time analytics and FHIR-enabled APIs.
- Replatformed legacy eligibility and provider systems to cloud-native architecture, cutting infrastructure cost by $2M.
- Built provider self-service portals and interoperability layers that reduced call-center volume by 48%.
Program Director at River Valley Behavioral Health (2015 – 2017)
- Oversaw 14 EHR implementations across Kentucky supporting 1,300 clinical users and led a 12-person team through recovery from a failed vendor implementation.
- Completed five go-lives in 18 months after a three-year stalled rollout by a previous contractor.
- Cut IT operating cost by $4M+ by renegotiating vendor contracts, eliminating redundant systems, and optimizing resourcing.
IT Manager at Luxottica (2011 – 2015)
- Directed global IT operations and data-center modernization for a $66B global retail brand, managing 145+ personnel across time zones.
- Delivered a $400M application and infrastructure migration from legacy hardware to a virtualized environment within 14 months.
- Reduced projected data-center operating cost by $130M over five years through consolidation and automation, and ran an offshore support center in China for 10K stores.
Links
Get your AI Cred
Costa earned this score in one conversation. The assessment takes 20 to 30 minutes. You talk through how you work with AI, and you leave with a score across five skills and a page of your own to share.
