AI-Powered Transformation for Sanofi’s Patient Support Programs

Coordinators spend 60–70% of their day on admin — not patients. Teqfocus deploys AI that eliminates the manual work in enrollment, prior auth, and patient engagement. Live in production in 10 weeks. No discovery phase.

×

About Teqfocus

The partner that owns the Salesforce layer and the data layer —and operates what it deploys

Most Agentforce implementations fail at the data layer. Teqfocus is one of the few partners that owns both — and one of fewer still that governs, tunes, and maintains agents after they go live. That’s not a feature. That’s the business model.

Market Positioning

Global Presence

The Opportunity

Dupixent PSP has the right ambition — and the right testbed to prove AI at scale

Sanofi is already investing in AI for PSP. The opportunity is to move from isolated experiments to production-grade AI that coordinators use every day — starting with Dupixent, then expanding across the portfolio.

18–25
day avg. enrollment-to-first-dose delay from manual intake and PA processing
60–70%
of coordinator time lost to data entry, status calls, and PA paperwork
3–5 days
from HCP referral to open case — from fax-based enrollment alone
59%
of patients unaware a support program exists for their therapy
Considerations as you scale AI across PSP workflows
01

Enrollment, intake, and prior auth are prime candidates for AI automation — and they compound each other

Manual SRF processing, fax-based referrals, and PA form-filling each add days to time-to-therapy independently. Automating them in sequence — intake feeds PA, PA feeds case management — produces compounding gains. These aren't separate projects; they're one connected workflow with one data model underneath.

02

Coordinator effectiveness is a data problem as much as a workflow problem

Without a unified CRM for PSP in Canada, coordinators make decisions without the full picture — which patients are at dropout risk, whose PA is about to expire, where the SLA clock is running. A Next-Best-Action engine that surfaces the right action for the right patient at the right moment transforms coordinator capacity without adding headcount.

03

Vendor-controlled data limits Sanofi's ability to act — integration is the prerequisite for AI

Hub vendors like McKesson control significant portions of the patient journey data. Layering AI on top of the existing stack — without replacing vendor relationships — requires a normalization layer that connects your data systems and payer APIs into a single actionable view. That integration architecture is where AI initiatives stall or succeed.

Teqfocus AI Demos

Four high-impact agents, ready to pilot

Each agent addresses a named pain point in Sanofi's PSP workflow. Each is a candidate for a 10-week pilot. Together, they form a complete AI-powered PSP operating model — deployable on your existing technology stack.

Priority 1 · Patient Journey

AI Next-Best-Action Engine — coordinator action queue, pre-drafted scripts

AI ranks patients by dropout risk and urgency across the full book of business. Coordinator sees the action, the reason, and a pre-drafted outreach script. One-click close. No separate data entry. AI recommends; coordinator decides.

Dropout Risk Model Pre-Drafted Scripts Human-in-the-Loop
30–40%
less coordinator decision
time per patient
15–20%
improvement in 6-month
patient persistence
50%+
of patients reached before
dropout signals peak
Priority 2 · Automate Routine Tasks

AI Intake & Enrollment Automation — same-day enrollment vs. 3–5 day lag

AI reads incoming SRFs — fax PDF, web form, HCP portal — extracts all fields, deduplicates against existing records, and pre-builds the case in the system of record. Coordinator reviews and approves in one click. Missing fields trigger an auto-message to the clinic. No manual keying.

SRF Extraction Deduplication Engine Auto-Resolution
50–70%
fewer manual touchpoints
2–3 hrs
freed per coordinator per day
Same day
enrollment vs. 3–5 day lag
Priority 2 · Automate Routine Tasks

Intelligent Prior Authorization — 10-day PA cycle compressed to 2–3

AI reads the patient's clinical profile, fills the correct payer-specific PA form, attaches supporting documents, and scores denial risk before submission. If denied, AI drafts the appeal letter from the clinical evidence library. Coordinator reviews and sends.

Auto-Populate PA Forms Denial Risk Scoring AI Appeal Drafting
10→2–3
day PA cycle
+15–20%
first-pass approval rate
+25%
appeal overturn rate
Priority 1 · Patient Journey

Personalized Patient Journey Engine — dynamic segmentation, every touchpoint adapts

Newly diagnosed patients get education-heavy onboarding and nurse visit prioritization. Experienced patients get proactive refill automation and minimal touchpoints. At-risk patients trigger intensive coordinator outreach. The AI adapts channel, frequency, content, and financial triggers per patient — in real time.

Dynamic Segmentation Dropout Prediction Multi-Channel Orchestration
Demo available on request
⚙️
AgentOps

Keep every agent performing after go-live

Built on your existing stack — no new infrastructure, no new vendor relationships.

Compliance — continuous PIPEDA and PHIPA posture monitored per agent, with on-demand audit reports cited to exact regulatory text

Agent performance — latency, token consumption, and response quality tracked per agent so degradation is caught before it reaches coordinators

Observability dashboards — agent health, utilization, and effectiveness surfaced in dashboards queryable by operations and compliance teams

Outcome accountability — Teqfocus owns the fix, not just the alert

Managed Services →

Compliance & Privacy

Patient privacy isn't a constraint we work around — it's a requirement we build to

Every agent Teqfocus deploys is designed for PIPEDA and PHIPA compliance from day one. Sanofi retains control at every layer.

Human-in-the-Loop AI

AI recommends; coordinators decide. No autonomous patient-facing actions without human review — by design, not by policy.

Data Stays Within Sanofi's Control

All data stays within Sanofi's environment. Nothing leaves to external AI training pipelines. Sanofi retains full data ownership.

De-identification Layer

Patient identifiers stripped before any model inference. PIPEDA and PHIPA compliant by design — not retrofitted after go-live.

Role-Based Access

Coordinator, manager, and field agent views are permission-gated. Full audit trail on every action, every interaction.

No Vendor Re-routing

Teqfocus layers on top of existing vendor relationships — McKesson, specialty pharmacies, payers. No new data flows without Sanofi's approval.

Release Gate with Sanofi

Sanofi retains approval authority on every production release. No surprise deployments. Biweekly steering cadence throughout.

Technology Partners

Layering on top of Sanofi's existing stack

Teqfocus integrates with Snowflake, Power BI, Coupa, and McKesson APIs. Your existing investments stay. We add the AI and Salesforce surface on top.

Salesforce
Salesforce
  • Summit Consulting Partner
  • 200+ Certified Experts
  • Health Cloud, Life Sciences Cloud, MuleSoft
  • 1,000+ healthcare customer success stories
Snowflake
Snowflake
  • Premier Services Partner
  • 20+ SnowPro Certified
  • 50+ customers
  • Healthcare-first data fabric accelerators
AWS
Databricks
  • Certified Consulting Partnert
  • 20+ AI & data workloads
  • MLOps pipelines
  • AI governance frameworks and model monitoring
Databricks
AWS
  • Advanced Consulting Partner
  • Data & Analytics Competency
  • 150+ active engagements
  • Cloud architecture for enterprise data platforms

Engagement Model

From kickoff to live AI — in 10 weeks

One use case. One dedicated pod. Live in production by week 10. No separate discovery phase — assessment happens in sprint week one.

1
Weeks 1–2

Org Access & Use Case Scoping

  • System access confirmed — read access to relevant data sources
  • Use case selected with Sanofi coordinators
  • AI model data requirements defined
  • Integration points mapped — Snowflake, McKesson, payer APIs
OutcomePilot scope agreed and signed off — build starts week 3, no ambiguity
2
Weeks 3–8

Build, Configure & Integrate

  • AI model built and configured on your technology stack
  • Flows, automations, and data connectors developed
  • Payer and McKesson API integrations live
  • UAT with Sanofi coordinators — edge case resolution
OutcomeAgent validated by the team who will use it — coordinator-approved before go-live
3
Weeks 9–10

Go-Live & Hypercare

  • AI capability live in production
  • KPI baseline captured — before vs. after tracked
  • Coordinator training and adoption support
  • Hypercare and AgentOps monitoring activated
OutcomeLive pilot with measured results — and a clear, data-backed case for what to scale next

Let's build Sanofi's AI-powered PSP.

A focused workshop to align on the highest-priority use case, confirm the data requirements, and set the 10-week clock. No slides. No discovery phase. A working session on what goes live first.