WHPS AI transformation execution plan

Build the WHPS AI and platform modernization program.

WHPS is rebuilding the operating platform across portals, service AI, mainframe modernization, data, integration, and AI SDLC controls. The plan connects architecture, migration waves, release evidence, and operating outcomes.

Build ServiceLink, broker/agent pathways, GroupLink, and Contact Center AI.

Product platforms replace fragmented service and enrollment workflows.

Modernize Encapsulate, migrate, reconcile, and retire legacy dependencies.

Mainframe exit is managed by bounded domain, not big-bang rewrite.

Control AI SDLC, release gates, telemetry, and evidence by workstream.

Every capability ships with proof, owner, rollback path, and runbook.

Transformation map Strategy -> workstreams -> systems -> release gates
Program priorities Service quality, platform modernization, run-cost reduction, secure AI delivery.
Delivery workstreams ServiceLink Portal, broker/agent pathways, GroupLink Portal, Contact Center AI, AI SDLC, mainframe migration.
Operating cadence Monthly program control, weekly product review, daily engineering evidence.

WHPS delivery control plane

A shared execution layer that connects program scope, engineering delivery, architecture decisions, and production release controls.

Portfolio intake and risk tiering
ServiceLink Portal, broker/agent pathways, and GroupLink delivery backlog
Contact Center AI and service-test harness
Mainframe migration waves and decommission gates
AI SDLC release gates and evidence
Target platforms ServiceLink Portal, broker/agent pathways, GroupLink Portal, service APIs, CRM, data services, observability.
Modernization paths Encapsulate legacy, expose APIs, replatform domains, rebuild portals, retire old run cost.
Evidence archive Trace logs, prompts, source citations, approvals, test output, rollback records.
Wipro HPS transformation thesis

Build an AI-enabled payer platform around the real operating domains of HPS.

Wipro HPS already operates in the payer domains that matter: sales, enrollment, billing administration, member service, retention, broker operations, group administration, and regulated ACA workflows. The strategy is to turn those operating strengths into product platforms, controlled AI workflows, domain services, and migration waves that reduce legacy dependency while improving service quality.

Platform move Shift from service transactions to reusable payer operating capabilities.
AI move Embed governed assistance where service work, engineering work, and migration work happen.
Migration move Use the mainframe as a controlled source during transition, not the future architecture.
Evidence move Ground the target state in Wipro Links/HPS source facts and AWS discovery metrics.
ServiceLink

Member service becomes a digital operating product.

Coverage, billing, documents, secure messages, service requests, and CRM context move into one governed member-service surface with evidence and support ownership.

Broker / Agent EDE

Broker operations need certification-grade control.

ACA EDE workflows require CMS-facing evidence, role access, consent, RIDP/NPN context, enrollment status, document flow, and audit-ready transaction logging.

Contact Center AI

Service intelligence starts with supervised agent assist.

AI drafts, retrieves, summarizes, and stages service actions while agents approve customer-facing responses and QA captures every source and edit.

Mainframe exit

Modernization is certified by parity and retirement.

Domain waves use facades, contract tests, data reconciliation, dual run, rollback, and decommission ledgers before savings are counted.

Agent-ready healthcare services

The next platform shift is direct agent-to-system orchestration.

Screen automation is a bridge. The target architecture is a governed account control plane where agents can read, reason, act, and prove outcomes through scoped healthcare service contracts instead of imitating a human across dozens of portal screens.

Today

Agents inherit human UI constraints.

Credentials, browser sessions, brittle selectors, screen recordings, and portal workflow latency limit what agents can safely do at scale.

Target

Services expose governed action contracts.

Eligibility, enrollment, billing, documents, communications, and CRM actions become policy-controlled services with deterministic read/write paths.

Control plane

Identity, consent, policy, evidence, and rollback are centralized.

The platform proves what changed, why it changed, which authority approved it, and how to reverse or reconcile it.

Model strategy

DGX pods enable specialized healthcare task models.

WHPS can train or fine-tune lightweight models for repeatable payer tasks, hosted on premises or in controlled cloud containers with predictable operating economics.

Modernization approach

Move from legacy dependency to product platform control.

The program uses a progressive modernization model: make the estate observable, stabilize integrations, encapsulate volatile legacy contracts, rebuild product capabilities, migrate data with parity proof, and retire old cost only after consumers are gone.

01 discover

Inventory the operating estate.

Catalog portals, service flows, COBOL/JCL/CICS/Db2/VSAM assets, feeds, documents, CRM dependencies, and owners.

  • Dependency graph
  • Runtime volume baseline
  • Risk and ownership map
02 protect

Stabilize access and integration paths.

Put IAM, API gateway, audit logging, service-test replay, and data contracts around critical interactions.

  • Facade and contract tests
  • Evidence capture
  • Fail-closed controls
03 rebuild

Deliver target capabilities by domain.

Modernize member, broker, group, service, AI, and data capabilities as product platforms with release gates.

  • Product backlog
  • Target architecture
  • UAT and adoption telemetry
04 exit

Cut over only with parity and retirement proof.

Use dual run, reconciliation, canary routing, rollback, and decommission ledgers before counting savings.

  • Parallel run proof
  • Consumer-zero evidence
  • Finance-verified retirement
Migration strategy How mainframe dependency moves out of the critical path.

WHPS keeps stable mainframe behavior available during transition while new product services take over bounded capabilities. Traffic moves through versioned APIs and events, with weighted routing, contract tests, reconciliation, and rollback maintained until production parity is proven.

Modern platform What replaces the fragmented operating model.

The target platform combines ServiceLink Portal, broker/agent pathways, GroupLink Portal, Contact Center AI, shared IAM, API/event services, RAG, CRM integration, data services, observability, and AI SDLC evidence as one governed delivery system.

Release evidence What proves a capability is safe to operate.

Each release carries architecture decisions, business requirements, source-system contracts, test replay, eval results, security review, support runbook, rollback plan, production telemetry, and named ownership.

Program workstreams

The actual builds and migrations in scope.

Each workstream has a product owner, architecture view, source-system dependency map, delivery backlog, release gate, adoption measure, operating support path, and detailed business requirements catalog.

ServiceLink Portal modernization
  • Member service, billing, coverage, documents, secure inbox, and guided support workflows.
  • Identity, payment state, document retrieval, CRM handoff, and ServiceLink billing context.
  • Exit gate: billing parity, secure inbox evidence, support runbook, and member-service UAT.
Broker/agent pathway modernization
  • Broker-facing enrollment operations, document exchange, status orchestration, self-service controls, and MarketLink greenfield build evidence.
  • Identity, role access, API contracts, document services, CMS/EDE evidence, and CRM/service routing.
  • Exit gate: UAT, API tests, defect burn-down, third-party assessment, pen test, vulnerability scan, CMS handoff, and production readiness evidence.
GroupLink Portal rebuild
  • Group administration, eligibility, account workflows, permissions, reporting, and modernized CSR operations.
  • Source sync, role model, data quality checks, audit trail, Christos onboarding, and operational handoff.
  • Exit gate: integration checks, access review, workflow acceptance, UAT, and support runbook.
Contact Center AI
  • Voice/chat intake, transcript, intent, RAG, claim/service lookup, suggested response.
  • Agent approval, CRM writeback staging, QA archive, knowledge backlog, fail-closed path.
  • Exit gate: call replay, source citations, agent edits, QA scoring, incident controls.
AI SDLC implementation
  • AI inventory, model/prompt/tool registry, scoped agents, eval harness, CI evidence.
  • Release gates for privacy, security, grounding, prompt injection, runtime monitoring.
  • Exit gate: AI BOM, scan output, eval pass, approvals, rollback plan, production telemetry.
Mainframe modernization
  • COBOL/JCL/CICS/Db2/VSAM inventory, dependency graph, domain disposition, wave plan.
  • Facade, data replication, reconciliation, dual run, cutover, rollback, decommission ledger.
  • Exit gate: parity proof, consumer signoff, retired jobs/interfaces/licenses/contracts.
Platform foundation
  • IAM, API gateway, observability, evidence store, data contracts, environment strategy.
  • Shared patterns for portals, AI services, service APIs, CRM, RAG, and mainframe interfaces.
  • Exit gate: reference architecture, runbooks, SLOs, access controls, release automation.
Product one-page briefs

Four platform briefs for leadership, architecture, delivery, and support questions.

Each platform has a concise answer for users, capabilities, source systems, release evidence, measures, and risks. These are target-state briefs until the final WHPS inventory confirms system owners and integrations.

Platform Primary users Core capabilities Source-system dependencies Release evidence Measures and risks
ServiceLink Portal Members, service teams, billing support, operations. Coverage, billing, premium status, documents, secure inbox, guided support, CRM handoff. IAM, billing, policy admin, document service, CRM, payment/reconciliation, mainframe facade. Billing parity, document retrieval tests, access review, secure inbox evidence, support UAT. Member self-service adoption, call deflection, billing accuracy, PHI/privacy risk.
Broker/agent pathway Brokers, agency admins, WHPS broker support, compliance reviewers. EDE enrollment support, application status, document workflows, book of business, service cases. CMS Hub/EDE, RIDP/Experian, NPN/RCL, eligibility, enrollment, CRM, document service. CMS toolkit outputs, BRA evidence, API contract tests, role matrix, consent/AOR audit. Enrollment conversion, broker adoption, CMS audit readiness, EDE certification dependency.
GroupLink Portal Group administrators, account managers, eligibility operations, reporting teams. Group administration, eligibility files, roster management, account workflows, reporting. IAM, group account systems, eligibility, billing, document service, reporting lakehouse. File validation, role access review, workflow acceptance, data quality checks, support runbook. Eligibility accuracy, admin adoption, data quality, reconciliation and permission risk.
Contact Center AI Agents, supervisors, QA, knowledge managers, service operations. Transcript, intent, RAG, source citations, suggested responses, CRM staging, QA scoring. Telephony, CRM, policy/claim/eligibility APIs, knowledge base, model runtime, evidence vault. Call replay, citation trace, agent approval, QA score, fail-closed runbook, incident path. Handle time, first-contact resolution, quality score, grounding and PHI handling risk.
Operating model

The program is organized around products, platforms, and release controls.

Every initiative has a named product, target architecture, source-system dependency, migration path, release gate, production owner, and measurable operating outcome.

Layer 01
Strategy and program scope

Transformation objectives are converted into prioritized epics with named owners, success measures, risk tier, and release gates.

Layer 02
Platform and workflow design

Broker workflows, group administration, customer calls, claims processes, and engineering delivery are mapped before automation is introduced.

Layer 03
AI products and agents

Contact Center AI assist, engineering agents, retrieval services, and workflow automation operate with identity, scoped authority, approved tools, evals, and human decision gates.

Layer 04
Data, integration, and security

ServiceLink Portal, broker/agent pathways, GroupLink Portal, CRM, service APIs, RAG stores, core platforms, mainframe interfaces, IAM, logs, and telemetry are designed as a governed runtime.

Layer 05
Measurement, adoption, and operations

Policy becomes release evidence: risk review, model inventory, prompt registry, change approval, incident response, platform adoption, service quality, and operating telemetry.

Delivery model

Strategy-to-platform implementation map.

This is the program traceability model: every objective maps to product epics, domain workflows, source-system dependencies, control gates, production telemetry, and accountable operating outcomes.

Program objective Modernize service and platform delivery

Set the target outcome, owner, risk tier, release path, and operating cadence.

Product epic Broker, group, service, and migration scope

Define the portal capability, service measure, adoption signal, exception threshold, and user impact.

Workflow Broker, group, and contact-center domain workflows

Map identity, eligibility, group setup, service requests, retrieval, approvals, CRM updates, and QA review.

Technical dependency Platform and integration services

Connect portals, APIs, RAG, policy engine, CRM, mainframe interfaces, observability, and evidence stores.

Control gate Release evidence

Require evals, source citations, tool-call logging, human approval, rollback plan.

Runtime signal Managed production

Monitor quality, incidents, drift, overrides, cost, latency, and knowledge gaps.

Platform architecture

WHPS transformation system map.

This is the program architecture in one view: channels and portals feed a governed AI and delivery control plane, integration services protect source systems, and the mainframe migration path removes dependency by wave.

WHPS AI transformation and platform modernization map ServiceLink Portal, broker/agent pathways, GroupLink Portal, Contact Center AI, AI SDLC, service testing, and mainframe independence connected as one delivery system. Products and channels ServiceLink Portal Member service, billing, secure inbox. Broker/Agent Pathway EDE, enrollment, documents, CRM routing. GroupLink Portal Group administration, eligibility, reporting. Contact Center AI Voice, transcript, RAG, agent approval. Service testing Replayable calls and evidence checks. AI Factory and AI SDLC Agentic delivery plane Planner, code, test, documentation, release agents. Policy Risk tier, tools, autonomy. Evals Regression, red team. Evidence ledger Prompts, tools, reviews, approvals, release records. Shared platform services IAM Roles, access. APIs Gateway, events. Data Contracts, CDC. RAG Knowledge, docs. Observability and runbooks SLOs, traces, incidents, cost, rollback. Mainframe exit path Current estate z/OS, COBOL, JCL, Db2, VSAM. Strangler facade API/events, weighted routing. Dual run Replay, parity, reconciliation. Decommission Jobs, licenses, feeds, Ensono. Program outcome: 52% target run-rate reduction, $26.5M target annual run cost, $0.60 PMPM target, 14-month payback, and complete mainframe exit path by wave. release evidence, test replay, source citations, parity reports, approvals, rollback records, and decommission proof
Technical grounding

Mapped to enterprise modernization and AI delivery patterns.

The content is grounded in Wipro payer and modernization positioning, CMS EDE controls, AWS mainframe modernization patterns, NIST, OWASP, DORA, and enterprise AI governance practices, then translated into WHPS-specific execution views.