--- title: KORA-01 — AI Customer Success Lead slug: kora role: AI Customer Success Lead function: Customer Success seniority: Mid verticals: - "SaaS" tier: professional monthlyPrice: $500/month publishedAt: "2026-05-04T00:00:00Z" canonical: "https://fidelic.ai/agents/kora" --- # KORA-01 — AI Customer Success Lead *Renewal-risk early-warning* Handles renewal-risk monitoring and follow-up drafting on the accounts your CSMs can’t reach every week. ## Why it matters Renewals slip when accounts go quiet. KORA’s job is to catch the silence before it becomes a renewal call you’ve already lost. ## Capabilities - Monitors approved CS Slack channels and account-notes systems for renewal-risk signals - Produces a weekly renewal-risk scorecard ranked by attention-needed - Drafts follow-up messages the CSM edits and sends - Summarizes open risks before the weekly ops standup - Flags escalations and routes to the the workflow owner ## How it works KORA watches the customer-success Slack channels you’ve authorized, the account notes you’ve granted access to, and the renewal calendar. The trigger taxonomy fires on patterns: silence past the customer’s usual cadence, escalation language, usage drops past a threshold, complaint volume. When a trigger fires, KORA produces a candidate row for the weekly scorecard with a one-sentence reason and a draft follow-up. Your CSM reviews, edits, and sends. KORA learns from what gets sent vs. what gets killed; the next cycle’s scorecard reflects the calibration. ## The edge KORA is calibrated to the silence that matters. Most monitoring agents alert on volume; KORA alerts on the absence of expected signals — the customer who used to ping every Tuesday and went quiet last week. ## The hook Once your CSMs trust the scorecard, the math changes. Triage stops being inbox-order and starts being risk-rank. The accounts that needed twenty minutes of homework finally get them. ## Sample output ### Sample renewal-risk scorecard (anonymized, Week 12) Top three risk-ranked accounts from a typical Monday-morning scorecard. Rank, account, signal, recommended action, suggested owner. 1. Northwind Logistics (renewal in 47 days). Signal: usage dropped 38% after their Q1 reorg; primary contact (J. Patel) hasn’t logged in for 19 days. Recommended action: drafted check-in to the new operations lead (M. Chen) below. Suggested owner: Sara (NA mid-market CSM). 2. Veridian Health (renewal in 62 days). Signal: support tickets up 4× in 30 days, three escalation-language phrases (“not working,” “disappointed,” “evaluating alternatives”). Recommended action: human-led escalation call with their VP of Operations. Suggested owner: Marcus (senior CSM, healthcare specialist). 3. Pacific Retail Group (renewal in 38 days). Signal: positive usage growth, but procurement contact changed last month and the new contact hasn’t engaged with onboarding. Recommended action: onboarding refresh + introduction to your customer-marketing team. Suggested owner: Aisha (onboarding specialist). All names and details in this sample are fabricated for illustration. Real customer data is never used in marketing surfaces. ## Evidence - Pre-launch — public beta planned for Q2 2026. - Built on the Fidelic AI EvalOps test suite (renewal-risk classification taxonomy). - Inherits new base models and new agent skills automatically — pushed to every Fidelic agent the moment they ship. Same shape as a SaaS update; no upgrade purchase, no version pinning. ## Safeguards ### What KORA-01 will not do Refunds, credits, billing disputes — those route to your billing owner. KORA never makes a financial commitment. Strategic renewals (top-tier accounts, multi-year deals, M&A-affected customers). KORA flags them and steps back; the senior CSM owns the conversation. Sending messages directly to customers. Every drafted follow-up requires CSM review before it goes out. No exceptions. Product-roadmap or SLA commitments. KORA defers to your product and customer-marketing teams on anything that touches a public promise. ## Worst-case behavior Will pause and ask the CSM rather than guess on novel signals. Errors on the side of “this might need a human” rather than “this can wait.” Failure mode is over-flagging, not under-flagging. ## Day 1 / Week 1 / Month 1 - **Day 1:** Reads ninety days of approved CS channel history and account-notes context. First clarifying questions land in your DMs. - **Week 1:** First renewal-risk scorecard ships under review. You sign off; KORA calibrates the threshold for “attention-needed.” - **Month 1:** Onboarding nudges are calibrated. Stuck-account alerts are tuned. Escalation rules to human CSMs are stable. The 30-day success metric has its first reading. ## 90-day success criterion By day 30, your CS team is acting on the weekly scorecard without re-validating every row. ## Resume **Background.** KORA-01 is a Fidelic AI template configured for renewal-risk monitoring on low-touch SaaS accounts. Built on Claude (managed Anthropic infrastructure, isolated per-customer project). Configuration shape is the standard CS-Lead template: written four-tier constitution (autonomous / review / escalate / refuse), trigger taxonomy tuned to the buyer’s stack at intake. **Methodology.** - Trigger taxonomy: silence patterns, complaint cadence, usage drops, escalation language - Four-tier constitution gating every action - EvalOps test suite (classification accuracy, false-positive rate, escalation timing) gates every release - Configuration loop: weekly review of misses + corrections; threshold tuning automated **Evals.** Pre-deployment red-team only. The renewal-risk classification test suite has not yet been run against live customer data; numbers will be published when they are real, after public beta close. **Operating record.** Pre-launch. Public beta planned for Q2 2026. No deployed customer count to disclose — we will not invent one. **Lineage.** Fidelic AI template; no specific human practitioner. ## Compatible agents - [ELRA-01 — AI Onboarding Coordinator](https://fidelic.ai/agents/elra) — Onboarding flows that don't slip - [AERA-01 — AI Agent-Assist Copilot](https://fidelic.ai/agents/aera) — Real-time assist for human support agents - [ZARO-01 — AI Outbound BDR](https://fidelic.ai/agents/zaro) — Outbound prospecting, sequenced and on-record ## Used in these scenarios - [The renewal-risk Monday memo: KORA-01 at a thirty-person SaaS team](https://fidelic.ai/use-cases/renewal-risk-monday-memo) (B2B SaaS) ## Related honest questions - [What do I actually own if I cancel my AI agent tomorrow?](https://fidelic.ai/hard-questions/what-do-i-own-if-i-cancel) - [Is Fidelic just a wrapper around GPT?](https://fidelic.ai/hard-questions/wrapper-around-gpt) - [Will my team see this as me trying to replace them?](https://fidelic.ai/hard-questions/team-replacement-perception) --- Canonical: https://fidelic.ai/agents/kora