“We’re not just building AI for call centers — we’re building AI with call centers in mind.” — Justin Massey, Relay Hawk

At the vCon Spring 2025 Conference, Justin Massey of Relay Hawk brought both a technologist’s vision and a contact center veteran’s intuition to the main stage. In a conversation with Technology Reseller News publisher Doug Green, Massey unpacked how his team is building next-generation AI voice agents — with human-first design and real-world BPO pain points at the forefront.

Contact Center Roots, Tech-Forward Vision

Massey isn’t your typical AI founder. With a family background spanning nearly five decades in call center operations, he was answering phones at 16 to fuel his Ford Explorer. “Some people theorize about customer experience. I lived it,” he shared.

That lived experience directly informs Relay Hawk’s core solution: AI agents that know when to get out of the way. When the AI hits a wall, Relay Hawk’s “escape hatch” hands the conversation off to a live agent — armed with the full context, so customers don’t have to repeat themselves.

AI Meets Authentication

One of the major challenges in today’s communication landscape? Verifying identity without collecting sensitive information. Massey proposes a modern approach: linking voice calls to identity providers like Google or Microsoft, much like how we authenticate via email or social platforms. “Why not use that same concept for phone calls?” he asked. Instead of collecting sensitive data during a call — a liability for BPOs — customers could validate via external credentials and receive a verified claim.

Tackling “Dependency Problems” in AI Integration

Relay Hawk is also focused on solving integration hurdles — what Massey called “dependency problems.” “Everyone wants their AI to plug into CRMs, ticketing platforms, schedulers — but every API is different,” he said. The goal: simplify third-party integration and leverage protocols like MCP to make AI deployment more seamless.

Warning: Transcription May Be Hazardous to Your AI

Another overlooked challenge: bad transcription = bad data = bad AI. Massey highlighted that many call recordings are mono, making it hard to distinguish between caller and agent. RelayHawk recommends stereo recordings, ensuring the data feeding AI engines is clean, clear, and actionable.

“If your transcription is off, your entire AI analysis goes off the rails.”

Relay Hawk Is Looking for Design Partners

Still in its early stages, Relay Hawk is working with design partners to refine its offering. “If you’re an early adopter and want to help shape a solution that actually works for your use case — we want to talk,” Massey said.

Learn more: RelayHawk.com or find them on LinkedIn by searching RelayHawk

#AIinCallCenters #VoiceAI #RelayHawk #vCon2025 #ContactCenterTech #IdentityValidation #TranscriptionAI #CXInnovation #BPOsolutions #CallCenterAutomation #ConversationalAI #DougGreen #TechnologyResellerNews