Salesforce’s 4,000 AI-driven layoffs and Oversee’s bid to try and automate 70% of TMC requests show how easily efficiency can come at people’s expense. Nathalie Lylock explains why Acai Travel takes a different path: deploying automation that strips away drudgery, upskills agents, and turns cost savings into richer customer experiences.

Watching the headlines—feeling the impact

Salesforce CEO, Marc Benioff, confirmed that 4,000 customer-support roles were eliminated because, in his words, “I need less heads” now that AI agents handle a growing share of cases. For every travel-industry leader who believes in technology’s promise, stories like this land with a thud. They remind us that behind every statistic is a human being with rent to pay, a career path, and a sense of purpose.

At nearly the same moment, an Israel-based reshopping platform, Oversee, announced ambitious targets: its Agentsee platform is targeting to auto-resolve 70 percent of inbound TMC emails and chats. Existing customers, like corporate travel agencies BCD Travel, and Christopherson, are piloting the solution, which Oversee says comes in “autopilot” (fully self-serve) and “copilot” (agent-assisted) modes.

Two news items, one unmistakable signal: travel operations are racing toward ever-higher automation thresholds. Yet the way we get there matters just as much as the percentage we automate.

Automation with empathy—my personal philosophy

At Acai Travel, we build AI agents that absorb the drudgery so people can amplify the distinctly human parts of service—judgment, creativity, and empathy. Our customers routinely shift up to 70 percent of repetitive work off their desks, but no one has had to announce large-scale layoffs because of our deployments. Why?

• Augment, don’t amputate.‍
  
  • We deploy AI in phases that give frontline teams time to redeploy talent toward higher-touch, higher-value interactions—irregular ops, VIP servicing, new market launches.

• Purpose-built guardrails.‍
  
  • Travel is uniquely complex—multi airline complex fare rules, interline agreements, NDC quirks. Our domain-specific architecture prevents “hallucinated” bookings and makes compliance audit-friendly from day one.

• Human-in-the-loop by design.
  
  • Every Acai Travel Agent includes escalation hooks, sentiment detection, and supervisor visibility so CX leaders can decide what should be automated versus what must stay human.
• Skills uplift, not skills atrophy.
  
  • We embed just-in-time knowledge panels that turn junior staff into mini-gurus, preserving institutional know-how instead of letting it walk out the door.

What Salesforce teaches us about change management

Salesforce’s Agentforce experiment proves that AI can slash ticket volumes. It also exposes the risk of chasing efficiency without an equally rigorous people plan. Customer-support organizations are ecosystems; you can’t simply remove half the biomass and expect the forest to flourish.

For travel brands, the stakes are even higher:

• Brand intimacy. A single disrupted trip can undo years of loyalty. Hand-offs between bots and humans must be seamless and emotionally intelligent.
• Regulatory complexity. DOT rules, EU261, and corporate travel policies create a labyrinth that generic AI often misreads.
• Ecosystem interdependence. Airlines, hotels, GDSs, and TMCs are stitched together by processes that break if one node goes dark overnight.
  

That’s why I always recommend a “three-lane” rollout:

1. Lane 1 – Inform & Assist. AI drafts responses, surfaces knowledge articles, but humans hit send.
2. Lane 2 – Execute & Verify. AI handles low-risk bookings, exchanges, and document retrieval with post-action human confirmation.
3. Lane 3 – Autopilot. Fully autonomous flows for standardized requests (invoice copies, simple point-to-point bookings) where data shows near-zero exception rates.
   

Only when each lane meets SLA and CSAT targets do we graduate more tasks—ensuring that workforce planning, upskilling, and morale stay in lockstep.

Healthy competition, healthier industry

I applaud Oversee for keeping us on our toes; competition forces all of us to innovate faster. But success won’t be measured only in deflection rates. The winners will be the platforms that:

• Protect traveler trust when disruptions hit hard.
• Free agents to be problem solvers, not process jockeys.
• Create new career pathways inside TMCs—think salespeople for complex VIP bookings, data-ops analysts, AI flow designers, CX strategists.
  

Those outcomes require more than code; they demand customer-centric change leadership. That’s the lane Acai has chosen.

A call to action for travel-ops leaders

1. Audit your ticket taxonomy. Identify which request types should be automated versus those where the human touch is irreplaceable.
2. Pair efficiency metrics with people metrics. Track agent engagement and skill development alongside AHT and cost per case.
3. Insist on transparent AI. Your team deserves to know why the bot offered a particular explanation or routed a PNR to a specific department.
4. Plan for growth, not shrinkage. Use labor saved to enter new markets, extend service hours, or deepen account management—turning cost savings into revenue growth.

Moving forward—together

AI is rewriting the rules of travel operations. Whether it writes a story of human displacement or human elevation is up to us. At Acai Travel, we choose elevation. We believe the future belongs to TMCs, OTAs, and airlines that harness automation to make every journey feel more personal, more resilient, and more human—precisely because the grunt work is gone.

If you’re ready to explore an AI roadmap that respects both your P&L and your people, my inbox is open.