Let's talk

Conversational AI, built to be trusted.

I'm an AI Product Manager and Conversation Designer. I led an enterprise migration from rule-based NLU (natural language understanding) to a multi-agent LLM system: 20M+ users, 8 markets, $2M in annual savings. I also run live production agents today, including the one in the corner of this page. Go ahead and talk to it.

Hiring a personSee my resumeFull-time roles, the one-page proof Hiring a firmWork with ICXConsulting, sprints, advisory

Christi Akinwumi, smiling, in a warm editorial headshot.

Numbers with receipts

Selected work

Case studies

All work
Case study · Side project Live demo

A 24/7 Booking Concierge for a DFW Photo-Booth Studio

A tool-using AI concierge that triages every lead, builds a real quote from a single pricing source, and lands a tentative hold on the actual Outlook calendar, at 11pm on a Tuesday, on autopilot.

<60s lead → quote + hold + SMS
  • Tool-use agent
  • Anthropic SDK
  • Microsoft Graph
  • Twilio
See it live Read the case study

Running live at boothlyevents.com.

Case study · Gated2026 Gated

Manual Guide Workflows → AI-Orchestrated Resolution

How manual lookups, disconnected systems, and poor handoffs turned a simple support issue into 24 minutes of customer frustration, and the architecture that collapses it to under 2.

<2 min90% time saved
  • Multi-agent
  • Orchestration
  • Handoff design
Read the teaser
Case study2025

Scaling LLM Support

Led an enterprise migration from rule-based NLU to an LLM multi-agent system with grounded RAG and AI output verification, across 8+ markets and 20M+ users.

25% faster$2M saved · 60% fewer hallucinations
  • Multi-agent
  • Hybrid RAG
  • LangFuse
  • Vendor evaluation
Read the case study
Case study2022

Self-Service Workflow Redesign

Rebuilt self-service for a connected-fitness product across 8 markets. IA, decision trees, and a product-gap loop that fed the roadmap.

40% faster35% adoption · 20% volume down
  • Workflow design
  • Information architecture
  • Localization
Read the case study
Case study2024

NLP Intent Taxonomy Redesign

Rebuilt the routing taxonomy from first principles. 150+ mutually exclusive intents across 12 product domains, derived from 50K+ real conversations.

30% misrouting down20% handle time down
  • NLP taxonomy
  • Intent classification
  • Conversation mining
Read the case study

About me

Hospitality taught me
to read the room.
AI taught me to scale it.

GED → hospitality → workflow optimization → conversational AI for 20M+ users.

More about me LinkedIn

Interactive · 2 minutes

What kind of AI collaborator are you?

Five honest questions, scored in your browser. Find out how you actually work with AI, and the one habit that would change your results.

Take the quiz

Interactive · 1 minute

A confident answer is not a correct one.

A conversation design teardown. Watch one support reply go from confident and wrong to calibrated, depending on what the assistant can actually know.

See the teardown

Writing

Essays

All essays
  • AI Evaluation
  • LLM Evals
  • Fundamentals

Evaluations are the product

The teams that deliver reliable AI are not the ones with the best models or the cleverest prompts. They are the ones whose evaluation harness got built first, committed to version control, and wired into every release.

May 2026·7 min readRead
  • Context Engineering
  • Prompt Engineering
  • Orchestration

Context Engineering: The Prompt Was Never the Point

The practitioners getting strong results from 2026-era models are not the best prompt writers. They are the best context architects. The real design artifact is what the system assembles around every prompt, turn after turn.

May 2026·8 min readRead

Work with me

Two ways to work with me.

Hiring a firmA few consulting engagements each quarter throughIntelligent CX Consulting . Start at services.

Hiring a personOpen to conversational AI, AI product, and applied AI roles. The resume is at /resume.