KT sells AI transformation to Korean enterprises while its own consumer, IPTV and broadband queues still run on IVR mathematics; a Gemini-powered agentic frontline deployed on KT's hardest service lines gives it a best-of-breed benchmark to run head-to-head against Mi:dm-based internal builds — and turns its own frontline into the proof point its AICT sales motion needs.
KT has publicly repositioned from telco to 'AICT' company, developing its own Mi:dm LLM family — with Mi:dm 2.0 released open-source in 2025 — and building AI contact-center products it sells to enterprises.
Public factKT signed a multi-year strategic partnership with Microsoft covering AI and cloud, including co-developing Korean-specialized models and a jointly-backed AI transformation business.
Public factKT has published AI-transformation revenue ambitions in the trillion-won range by the late 2020s, making AI delivery a board-level commitment.
Public factKT's consumer business spans mobile, broadband, IPTV and fixed-line — bundle billing and subscription-change volume with legacy-system complexity underneath.
Reasoned inferenceKT's own AICC products create channel conflict: it sells contact-center AI while buying it — any external platform must position as benchmark and acceleration, not replacement.
Reasoned inferenceMedia/IPTV service lines (set-top troubleshooting, bundle changes) likely carry the worst handle times and the oldest customer demographics.
Seller hypothesis — validateValidate with the account team before outreach: Status and scope of internal Mi:dm-based contact-center deployments on KT's own lines · Microsoft-alliance exclusivity terms as they touch conversational-AI vendors · Which service lines are open to external platforms versus reserved for internal builds · Benchmark metrics KT would accept as decisive
Strong internal engineering — sell infrastructure, models, governance, or selected capabilities
Evidence: KT is a genuine builder with named public evidence — the Mi:dm LLM family (Mi:dm 2.0 released open-source), AICC products it sells commercially, and a strategic Microsoft alliance with published AI-transformation revenue targets — so the platform sale is a co-build motion: benchmarkable deployments on selected lines alongside KT's own stack, never a platform land-grab.
Why they won't build the full stack: Even a committed builder faces a sequencing problem: consumer queues need relief on this year's cost curve while Mi:dm-based AICC capabilities mature, and an external best-of-breed benchmark makes the internal program measurably better — the pitch is acceleration and evidence, not substitution.
KT has publicly committed to its AICT transformation — Mi:dm LLM development, a multi-trillion-won Microsoft partnership and stated ambitions to grow AI-transformation revenue several-fold by the late 2020s.
GTM implication: Frame every proposal inside KT's own declared strategy: a benchmarkable frontline deployment is evidence for the AICT story, not a challenge to it.
KT's investor communications emphasize cost efficiency and returns discipline alongside the AI investment program.
GTM implication: The pilot's cost-per-contact math funds the AI narrative — position it as the efficiency proof the investment program needs to show.
| Workflow | Why it matters here | Value | Complexity | Speed | Channels |
|---|---|---|---|---|---|
Converged billing and subscription service line Bundle billing and plan changes resolved in-conversation with transactions executed; a measured benchmark against internal-build economics. | Bundle billing across mobile, broadband and IPTV is KT's structural volume — high-frequency, legacy-complex, cost-heavy. Friction today: Bundle questions span systems; plan-change journeys break between app and call; billing-cycle peaks melt queues. | VoiceApp chatKakaoTalk | |||
IPTV and broadband technical triage Structured diagnosis with account and line data, guided fixes in-flow, truck rolls reserved for genuine hardware failures. | Set-top and connectivity issues drive long handle times and truck rolls that a diagnosing agent can often prevent. Friction today: Scripted human triage misdiagnoses; escalations queue; truck rolls dispatch for issues resolvable in-conversation. | VoiceApp chat | |||
Retention and bundle-renewal desk Full expiring-cohort coverage with policy-bounded renewal offers; save economics measured end-to-end. | Contract-expiry cohorts shop the three-carrier market; proactive renewal conversations at full coverage beat dialer-limited saves. Friction today: Save teams reach a fraction of expiring contracts; offers vary by agent. | KakaoTalkVoiceSMS |
Voice & channel orchestration, telephony, conversational execution, session/state, routing, integration build. Capability coverage validated during implementation.
API access to these systems is the critical-path dependency.
Identity-bound sessions, policy-bounded actions, 100% audit logging, human approvals at defined points, in-tenant intelligence.
All figures are modelling estimates from the labeled inputs above — nothing here is customer-provided yet. The pilot's first job is replacing these assumptions with KT Corporation's measured baseline. Package price covers implementation only; recurring usage billed separately.
Why this package for KT Corporation: A focused pilot on one high-volume line fits an account that will insist on benchmarking against internal Mi:dm-based builds before any platform commitment — win the bake-off first.
Packages cover implementation and integration only. Recurring costs are billed separately: Tilicho Labs platform usage (~$0.15/call-min indicative, usage only), Google Cloud consumption, telephony/carrier charges, managed operations, and support & optimization. No package includes unlimited usage.
“KT sells AI transformation to Korea; its own service queues are the most public test of whether AICT is real. A benchmarked best-of-breed frontline de-risks the story you have already told the market.”
“Run this as a controlled bake-off: Gemini-powered platform on one line, Mi:dm-based build on another, same metrics — you keep architectural sovereignty and gain the external benchmark that sharpens the internal roadmap.”
“Bundle-billing peaks and IPTV triage carry your worst handle times; elastic capacity on those lines cuts cost now while the internal platform matures on its own timeline.”
“Your queues cannot wait for the AICT roadmap; a production pilot delivers relief this quarter and the benchmark data that makes every internal-build decision smarter.”
“PIPA-aligned handling, 100% logging and script governance on a production line — external-platform governance your internal builds can adopt as the standard.”
“A Launch-scale pilot prices the benchmark at a rounding error against the AICT investment program — and produces the cost-per-contact evidence that disciplines the whole portfolio.”
Written from this account's own research — the strategic signal, the capability gap, the entry workflow, and the modelled economics — not a mail-merge template. Pick the moment and the persona, edit anything, then copy or open in your mail client.
Customer-safe by construction: drafts are composed only from customer-facing fields. Account tier, build-vs-buy classification, priority score, internal routing, and partner-commercial detail are not inputs to the composer, so they cannot appear in a draft. Money figures are always framed as modelled from the customer's own volumes. Read before sending — you own what goes out.
Korea's #2 telco has publicly bet the company on an 'AICT' transformation — its own Mi:dm LLM family, a multi-trillion-won Microsoft alliance and a declared AI-transformation revenue target — making it simultaneously Korea's most AI-committed carrier and a builder that will benchmark any vendor hard.
Benchmark-framed pilot proposal to the AX and customer-service organizations: one high-volume line, shared metrics with any internal build, results reviewed jointly in a quarter.
Entry: Converged billing and subscription service line (mobile, broadband, IPTV bundles) · Launch package · 8–10 weeks to a live, measured pilot. Human fallback throughout; success thresholds agreed before build.
Start the pursuitResearch-based priority-account universe assembled from public information, market scale, communication volume, and solution fit. This is NOT an authoritative list of top Google Cloud customers; existing Google Cloud relationships are noted only where publicly reported. Validate every account with the account team before outreach.