Senior Manager, Support Operations (AI & BPO)

Amira Learning

29d•Remote

About The Position

Amira Learning accelerates literacy outcomes by delivering the latest reading and neuroscience with AI. As the leader in third-generation edtech, Amira listens to students read out loud, assesses mastery, helps teachers supplement instruction, and delivers 1:1 tutoring. Validated by independent university and SEA efficacy research, Amira is the only AI literacy platform proven to achieve gains surpassing 1:1 human tutoring. Trusted by more than 2,000 districts and working in partnership with twelve state education agencies, Amira is helping 3.5 million students worldwide become motivated and masterful readers. Amira's Support function is small, the BPO partner is mid-onboarding, and our chatbot needs work. We are hiring a Senior Manager, Support Operations to own end-to-end frontline support — Tier 0 (our AI chatbot) and Tier 1 (our BPO team) — and to operate the whole thing as an AI-native function. We are not hiring a BPO manager who is curious about AI. We are hiring someone for whom AI is the substance of the work — writing prompts, building agent-assist tooling, analyzing tickets at scale, integrating systems with MCP, and continuously improving how our chatbot deflects and resolves. The BPO team is one of the things they operate, alongside and through these AI surfaces. The right person treats every operational question — staffing, QA, escalations, knowledge gaps — as something to investigate with data and improve with code, not something to manage through spreadsheets and meetings.

Requirements

You have built and shipped AI-powered support tooling that real people use — agent-assist features, LLM-powered ticket analysis, chatbot improvements, or similar. We will ask you to walk us through what you built and how.
You have hands-on experience writing prompts as a working practice — for production systems, not personal experiments
You have integrated AI tools with operational systems (CRM, helpdesk, data warehouse) via MCP, APIs, or equivalent
You have worked in a customer support or customer experience function — running queues, owning SLAs, working with vendor teams, or analyzing tickets at scale
You are comfortable in code: reading it, writing it with AI assistance, debugging it when it breaks. No specific language requirement.
You can have a direct conversation with a frustrated customer or district administrator without losing your composure

Nice To Haves

This is a builder's role. Expect to spend significant time: In Claude Code, Cursor, Copilot, Codex, or equivalent — writing prompts, agent flows, integration code, analysis scripts. Daily, not occasionally.
With AI coding agents as a working partner, not a curiosity
Reading tickets, transcripts, and customer conversations — at scale, with AI assistance, looking for patterns no spreadsheet would catch
Building agent-assist tooling and MCP connectors — and shipping them iteratively
In Salesforce, Zendesk, the chatbot platform, and our data layer — comfortable enough to investigate, query, and modify directly
If you've never written a prompt, configured an MCP connector, or sketched an agent flow, this role will be a hard ramp. If you do these things daily already, you'll feel at home.

Responsibilities

Continuously improve our customer-facing chatbot: prompt engineering, knowledge base curation, intent and entity tuning, escalation logic
Track and improve containment rate, deflection rate, and CSAT for chatbot-handled interactions
Build the diagnostic trees, agent flows, and fallback handling that make the chatbot trustworthy at scale
Iterate weekly, not quarterly
Own the day-to-day relationship with our BPO partner, including agent-level performance management, QA calibration, and SLA accountability
Run the operational rhythm: weekly syncs, monthly reviews, quarterly business reviews
Build agent-assist tooling (LLM-powered response drafting, knowledge retrieval, ticket categorization, escalation routing) that makes our BPO agents materially faster and more accurate
Drive coaching and quality by using AI to analyze 100% of tickets — not 10% samples
Build and maintain MCP connectors wiring AI tools into our operational systems (Salesforce, Zendesk, ChurnZero) so AI agents can read and act on real data
Personalize agent-assist responses with customer context pulled from CRM and product data
Design integrations that respect FERPA obligations and the sensitivity of educator and student data
Run AI-driven analysis of ticket volume, contact drivers, sentiment, and resolution patterns
Translate analysis into concrete improvements: knowledge base updates, chatbot tuning, training topics for BPO agents, product feedback for Engineering
Build reusable dashboards and reporting that make support performance legible to the rest of the company
Apply FERPA and COPPA rigor to every AI system you build — guardrails, audit logs, human-in-the-loop where student data is in play
Own agent background-check and approval workflows for BPO personnel with access to student PII