Ant International brings AI-powered customer service to Asia
Ant International is partnering with TNG Digital and easypaisa to roll out generative AI for customer support across e-wallet and digital banking. The goal is straightforward: faster responses, fewer manual steps, stronger security, and support teams that ramp up in less time.
What this means for customer support leaders
TNG eWallet and easypaisa are using the Alipay+ GenAI Cockpit to automate responses and streamline how requests are handled. The AI agents use domain knowledge to personalise help in-app and complete routine tasks without human handoffs.
For support teams, this can cut training time, reduce handling time on common issues, and keep service quality more consistent. It also creates a clear path to scale support without scaling headcount at the same rate.
Personalisation and security in one flow
easypaisa is adding more personalised product recommendations, financial insights, and reminders-directly inside the app. Alongside this, the system adds advanced risk detection to keep sessions safer while customers get advice that fits their situation.
"Our strategic collaboration with Ant International is democratising access to easy and convenient digital financial services for millions nationwide. The launch of our generative AI-powered customer service agent is a key step that enables easypaisa customers to enjoy truly hyper-personalised financial experiences, from tailored product recommendations and actionable insights, to timely reminders that help them manage their finances more efficiently. Beyond personalisation, the AI assistant enhances security through advanced fraud detection and improves risk management, making every in-app financial interaction safer, faster, and more intuitive. This innovation represents a significant leap in making digital financial services more accessible, trustworthy, and empowering for all Pakistanis," said Farhan Hassan, Chief Digital Officer, easypaisa Digital Bank.
Inside the Alipay+ GenAI Cockpit
Launched in 2025, the Cockpit draws on 20+ large language models, including Ant International's proprietary Falcon TST (Time-Series Transformer). It connects to banking-specific datasets-like transfer protocols and dispute policies-so responses reflect actual rules and workflows.
The platform supports copilots and specialised agents for fintech operations. Example: Antom Copilot helps merchants integrate payment solutions more efficiently, taking pressure off human teams for setup and troubleshooting.
Security built for fintech support
Ant International's AI SHIELD toolkit is embedded for compliance and risk control. Its AI Security Docker targets up to a 90% reduction in AI service risks through security assessments, real-time monitoring of agent activity, and continuous testing.
If you're building guardrails of your own, frameworks like the NIST AI Risk Management Framework can help standardise your approach to oversight, logging, and escalation.
How to integrate AI support without breaking your queue
- Start with high-volume use cases: account access, top-ups, failed transfers, chargebacks, KYC/verification status, and dispute triage.
- Centralise knowledge: policies, fees, SLAs, eligibility rules, and exception paths. Keep content versioned and auditable.
- Define strict escalation rules: confidence thresholds, sensitive intents (fraud, account lock, chargeback disputes), and regulator-triggered flows.
- Wire into existing systems: CRM, ticketing, IVR/chat, fraud/risk engines, and payments backends for action-taking (with role limits).
- Set guardrails: PII handling, retention windows, model access scopes, rate limits, and blocked intents.
- Monitor the right metrics: deflection rate, AHT, FCR, CSAT, containment on sensitive flows, time-to-proficiency for new agents.
- Run continuous QA: red-team prompts, hallucination checks, policy drift tests, and weekly review of sampled conversations.
- Train the team: prompt patterns, handoff etiquette, compliance basics, and when to override the bot.
Why this is gaining traction now
Support teams in fintech face rising volumes, strict compliance, and customers who want instant answers. A multi-model approach that's grounded in real banking rules gives you faster resolution without cutting corners on risk.
"We're excited to welcome our first customers to Alipay+ GenAI Cockpit. This platform opens the door for fintechs to agentic AI - empowering them to automate decisions, accelerate innovation, and deliver smarter financial solutions to their partners, all with financial-grade compliance, reliability and security. We look forward to supporting our customers as they expand the use of the Cockpit platform in their operations," said Jiang-Ming Yang, Chief Innovation Officer at Ant International.
Next steps for your support org
- Audit your top 50 intents and tag them by risk level, automation fit, and required data sources.
- Pilot one secure, high-impact flow (e.g., failed transfer resolution) before scaling to disputes or account changes.
- Stand up a weekly AI Ops review: metrics, error analysis, content updates, and policy compliance checks.
If you're upskilling your support team for GenAI operations, see curated training by role at Complete AI Training.
Your membership also unlocks:
