Who Is Leading the Build
Harvey Hongwei Li, PhD
Founding AI/ML engineering team at both Uber and Airbnb. Built and scaled the machine learning infrastructure that powered core platform intelligence at both companies during their highest-growth phases. PhD from UC Berkeley.
A builder and innovator who believes the next great infrastructure problem to solve is not another consumer platform – it is the $2 trillion healthcare labor market where 60% of every dollar goes to workforce and the systems that manage it have never been connected.
Founding AI/ML Team
Founding AI/ML TeamPhD, UC Berkeley
San Francisco
What Harvey Is Building
Harvey is building the technology for a new category: Workforce Unifying Infrastructure. An AI-native platform that sits between every system of record a health system runs, fills the cracks between them, and turns fragmented data into unified intelligence that agents can act on. Not another HR tool. Not another dashboard. The infrastructure layer that has never existed for the $2 trillion healthcare labor market.
Who It's For
Every executive sees a different problem. We solve all of them.
"AI is disrupting every workforce function. Labor is 60% of costs. Skills and tasks change faster than we can track. First-year attrition over 45%. The board wants overhead reduced."
Unified intelligence across skills AND tasks. Real-time visibility, career pathways, and AI agents that amplify your team.
See gaps before they become vacancies
The Problems
Name what's broken.
10+
siloed systems
Workforce Data Fragmentation
10+ disconnected systems per health system. None talk to each other.
Administrative Workforce Infrastructure
High costs are only half the story. Fragmented admin infrastructure destroys productivity across every function it touches.
Worki amplifies productivity across every function – not by eliminating people, but by amplifying them.
The Solutions
How we fill the cracks and strengthen the whole.
One architecture. Four tiers. Named agents governed by humans. Shared infrastructure that compounds with every member. Here is how it works.
Workforce Unifying Infrastructure (AI)™
The Architecture
AI Models
Claude | GPT | Gemini | Local Models
▼ORCHESTRATES
Multi-step workflows with human approval gates. Agents execute across credentialing, onboarding, and redeployment with Human Conductor at every decision point.
Agent Orchestration
Amplification + Research Agents | Human Conductor™
▼REASONS OVER
Intelligence layer provides context for every agent decision. Skills, tasks, credentials, career pathways, and risk signals inform what agents do and when humans intervene.
Unified Intelligence
Skills | Tasks | Risk | Pathways | External Data | Crosswalk
▼NORMALIZES FROM
Raw data from disconnected systems into a unified skills-task-credential schema. Flat files, APIs, scraped public data, and coalition-shared intelligence all flow through.
Integration
Licensed Agent Standard™ | Flat Files | APIs | Between the Seams
▼CONNECTS TO
Pipeline integration between the seams of existing systems. Licensed Agents login under governed credentials. No rip-and-replace.
Systems of Record
HRIS | Credentialing | Scheduling | Learning | Payroll | ERP
Each layer depends on the one below it. The strength lies between.
Four tiers. One infrastructure.
Click each tier to explore.
1
Unify™
Five intelligence layers fused in real time.
▼
2
Pathways™
Scaffolding, mobility, tax-advantaged upskilling.
▼
3
Amplifiers™
Named agents with Human Conductor™ governance.
▼
4
Infrasharing™
Shared intelligence. Coalition economics. You choose.
▼
CIO
Infrastructure Deployment
The questions every CIO and executive team asks. Answered head-on.
We just spent $50M on Oracle/Workday. You want us to replace it?
No. Worki sits between the seams of your existing systems. We aggregate – we do not replace. Licensed Agents login to your systems under governed credentials with full audit trails. Your investment is preserved.
Every AI vendor promises easy integration. It always takes 12 months.
We start with flat files. CSV exports from your existing systems. No API needed. Build intelligence in weeks, prove value before any integration project begins. APIs come when you are ready – on your timeline.
How do Licensed Agents work? What is different from another SaaS login?
Licensed Agents carry their own security certification, audit trail, role-based access boundaries, and compliance documentation. Think of them as pre-certified team members that meet your enterprise security standards from day one – not a platform your team has to learn.
What about data security? Patient data? HIPAA?
Worki operates on workforce data – not clinical or patient data. For coalition participants, intelligence and insights move across the network, but raw data never leaves your governance perimeter. Data sovereignty preserved per system.
Our IT team is stretched thin. What is the lift?
Phase 1 requires zero IT integration. Flat file ingestion runs outside your systems. As you expand, Licensed Agents connect via standard APIs that your team already manages for Oracle, Workday, or UKG. Minimal net-new burden.
What if we want to start small?
That is the architecture. Start with one role, one department, one agent. Prove value in 60–90 days. Expand as capacity and confidence grow. Modular Role Architecture™ means agents deploy role-by-role, not as a monolithic launch.
Doesn't Workday or UKG already do this?
Workday, UKG, Oracle, and ADP are excellent systems of record. They provide tremendous value and have earned their place in every health system. But they are designed to store and manage data within their own environments. Connecting intelligence across systems, deploying governed agents that operate between platforms, and sharing workforce intelligence across institutions would fundamentally change their business model. They are the systems. We are the connective tissue between them. Worki is system-agnostic by design – we make every system of record more valuable, not less.
Beyond Healthcare
Healthcare first. Then every complex industry facing the same gap.
The workforce fragmentation problem exists across every complex enterprise industry managing large, credentialed, distributed workforces through disconnected systems. The architecture is the same. The intelligence compounds across verticals.
🧪
Pharmaceutical / Life Sciences
Thousands of credentialed researchers and manufacturing specialists across fragmented compliance, training, and credentialing systems. FDA audit trails and GMP certifications managed manually across sites and CROs. A global pharma company with 30,000+ employees across 50 sites has no unified view of workforce readiness for FDA inspections.
⚡
Energy
Complex regulatory environments, distributed field operations, and safety-critical credentialing across nuclear, utility, and renewable workforces. Certifications expire, safety training lapses, and workforce planning relies on spreadsheets. A regional utility with 12,000 employees runs separate credentialing systems for every operating division.
📡
Telecom
Rapid technology transitions (5G, fiber, AI-driven networks) are making entire workforce segments obsolete. Tens of thousands face role elimination as automation accelerates. No infrastructure to map exiting workers to emerging AI-enabled roles. A Fortune 50 telecom with 90,000+ employees where 15,000 legacy roles face elimination over 36 months.
🏦
Financial Services
Regulatory credentialing (Series licenses, compliance certifications), complex structures spanning retail banking through investment operations. Manual compliance tracking across business units creates audit risk. A diversified firm with 25,000+ employees across banking, insurance, and wealth management – each division running separate compliance and credentialing systems.
Start small or start large. Phase in. Scale up.
The human leads, the agents amplify, and the infrastructure compounds.
Request a Demo
Read the Book →