Case Studies / Medtronic

Transforming Patient & Clinician Support from Conversational IVR to Agentic AI

Discover how Medtronic modernized patient and clinician experiences across voice and digital channels with Teneo—evolving from early deterministic virtual assistants to secure, compliant agentic AI.

$6M Cost Saved • +36000 Agent Hours Saved • -37% Wait time • -55% Misrouted Calls

Orchestration First

The rapid deployment allowed Medtronic to combine deterministic flows + LLMs + tools with guardrails.

Voice-grade NLU

Purpose‑built for IVR and noisy real‑world speech (e.g., “pacemaker”≈ “pizza maker” correction patterns).

Built‑in Compliance

The Teneo Platform is built for the enterprise with SSO, PII scrubbing, policy routing, and auditable citations.

With Teneo, Medtronic delivers faster resolutions, materially lower costs, and the governance needed for regulated healthcare across 10+ business units and 60+ contact centers, supporting 2,000+ agents

The Results

Medtronic is the leading global healthcare technology leader with solutions that assist with over 70+ health conditions. These innovations span groundbreaking and lifesaving medical technology, such as cardiac devices, surgical robotics, insulin pumps, surgical tools, and patient monitoring systems.

Medtronic modernized patient and clinician experiences across voice and digital channels with Teneo—evolving from early deterministic virtual assistants to secure, compliant agentic AI. Following the launch of Teneo Agentic AI, the results have been remarkable:

Cost & Scale
- $6M costs saved (2022) and 36,000 agent hours saved
- Cost per contact down from $25.96 to - Program supports ~700 agents in the contact center
- Estimated $9–10M cumulative savings in the Cardiovascular Group business unit to date
- 45% YoY demand growth absorbed via self‑service/virtual agents.

Field Productivity & Revenue Impact
- Cranial & Spinal: 4–5 hours/week given back to each field rep—enabling more sales calls and revenue capture.

Efficiency & Experience
-37% wait time vs. pre‑Teneo
-55% misrouted calls (9% → 4%)
-6.7 pts abandonment (14.7% → 6.8%)
+18 pts service level (51% → 69%)
+6% CSAT (and +2% in the initial year)
>1.05M voice IVA sessions handled (by Jun ’23).

Teneo Agents Live
(17 new agents in the roadmap)

6M

Cost Saved

1.05M

Calls Handled

+36000

Agent Hours Saved

+8%

Customer Satisfaction

-37%

Wait time

-6.7%

Abandonment rate

+18%

Service Level

-55%

Misrouted calls

Download this Case Study

Get the complete analysis of how Medtronic achieved unprecedented ROI with AI contact center automation. Includes detailed metrics, implementation roadmap, and strategic insights.

Michael Altieri of Medtronic — **Michael Altieri**
Service Delivery Manager, Virtual Assistants at Medtronic

The Challenge

Previously, Medtronic supported 10 business units with 63 call centers that employed over 2000 call agents, with each department supporting specialized types of enquiries depending on patients’ medical issues and the kind of medical device they need information on. The departments were also the point of contact for medical professionals who wanted to make new purchases or find information on specific products. However, the performance and handling of calls was not at an optimal level.

Many of these problems were caused by an old keypad IVR system that failed to provide a suitable level of care for customers and medical patients. Furthermore, the staff were negatively affected and overwhelmed by the impact of misrouting and basic troubleshooting taking up vast amounts of their time.

Critically, these issues were being recorded in a unit focused on providing information on cardiovascular products, where patients needed information on the performance of various products such as pacemakers and defibrillators. The speed at which calls were answered and routed was of critical importance but due to the poor performance of the legacy IVR system, the company was not able to provide the service level or experience required for its customers.

Legacy keypad IVR created misroutes, long wait times, and inconsistent experiences in high‑stakes cardiovascular support. Agents lost time to basic troubleshooting and manual lookups, while strict industry requirements demanded strong security and governance.

The Solution

2018–2019: Foundations: Began with deterministic assistants; first Virtual Agent live Aug 2019. Internal IT assistant “Sparky” launched.2022: Conversational IVR Pilot: Implemented Teneo Conversational IVR to route callers (patients/clinicians/sales) to the right expertise. Integrated Five9 for callbacks and SMS with links to approved video instructions and step‑by‑steps.2023: Scale & Advanced Troubleshooting: Expanded Teneo to handle all patients and clinicians calling in. Introduced advanced troubleshooting: when a device monitor alarmed (e.g., pacemaker), Teneo captured caller input and delivered stepwise, approved guidance or SMS’d the right content. Adopted RAG on approved content with transparent citations (“Don’t trust the model—trust the citation”), plus post‑processing guardrails and handoffs between assistants (e.g., general → CVG).2024–2025: Toward Agentic AI: Security by design: completed a 7‑month vendor risk assessment with findings remediated; HIPAA/PCI‑compliant deployments; PII scrubbing/anonymization of logs. PagerRep workflow agent launched to page the right field rep digitally, replacing phone‑only flows. Employee SSO for protected assistants; routing orchestration across domains.Future: Extend proactive care with Virtual Earl so it can reach out to patients based on medically relevant device events (with governance/human‑in‑loop where required). Expand agentic patterns across business units with multi‑agent planning and tool use. Continue model diversification and advanced analytics on transcripts.

Learn more about AI voice agent for healthcare.

The Power of
Teneo

We help high-growth companies like Telefónica, HelloFresh and Swisscom find new opportunities with our technology.

Interested to learn what we can do for your business?

Get in Touch