For back-office and shared-services leads

AI for Back-Office and Shared-Services Operations

Rules-heavy, high-volume work with real error costs. The right automation pays for itself fast. The wrong one creates an incident. I help you tell the difference and ship the right one.

Book a 30-minute callTake the 2-minute scorecard

No pitch. If I can’t find a workflow worth automating, I’ll tell you on the call.

The work this covers

Identity and employment verification. Customer and vendor onboarding. Billing disputes and adjustments. Document intake and review. Order exceptions and fulfillment follow-up. Internal requests that land in a shared inbox. The common shape: a case comes in, someone checks it against rules and records, and a decision or a document goes out.

That shape is where AI delivers its clearest returns, and where it does the most damage when it is wrong. A mistaken verification can cost someone a job or a lease. A bad billing adjustment is a refund and a churned account. So the design question is never just "can we automate this" but "what has to stay with a person, and how do they catch what the model misses."

How I approach it

I start by pulling the case data and sitting with the people who work the queue. We sort the work into three buckets: cases that should not exist (fix the upstream cause), cases that are routine and well defined (automate with review), and cases where judgment or exposure is high (keep with people, give them better tools). Then we baseline the routine bucket and build the first automation with a human checking every output until the error rate earns trust.

  • Case classification and routing so the right work reaches the right person first
  • Extraction and pre-fill from documents, with confidence thresholds and review queues
  • Decision support that drafts the recommendation and cites the records, rather than deciding alone
  • QA sampling that scales with volume instead of headcount
  • Audit trails designed in, so compliance can see what the model did and why

Why I am the right person for this

I lead strategic operations at Checkr, where automated screening, accuracy, and FCRA compliance have to coexist on every case. Before that I ran the support and tooling platforms behind 500,000 monthly contacts at Lyft, including the migrations that moved twelve sites onto new systems. I have owned the budget, bought the platforms, and been the person accountable when a rollout touched real customers.

Questions

What people ask first.

How do you handle compliance and audit requirements?

As infrastructure, not an afterthought. Every automated step logs its inputs, its output, and who reviewed it. Thresholds for human review are set with compliance in the room, and the first weeks run with full review so the error profile is known before anything is trusted.

Our data lives in three systems and a lot of email. Is that a blocker?

It is the normal starting point. The assessment maps where the work actually lives, and the first project usually includes getting one workflow into a system where it can be measured. That alone often pays for itself.

Which AI tools and vendors do you work with?

I am vendor neutral. I have bought and implemented platforms such as Zendesk, Twilio Flex, and 8x8 at Lyft scale, and I build with Claude, OpenAI, and current agent tooling every week. The recommendation follows the workflow, not the other way around.

Do we need engineers for this to work?

Usually not for the first wins. Most operations automation comes from well-defined workflows and off-the-shelf tooling configured carefully. When a build is needed, I write the first version or a spec your engineers can act on.

We are in a regulated industry. Is that a problem?

It is the normal case for me. I run operations under FCRA today, and at Lyft I worked through regulatory programs that required careful tradeoffs. Compliance is designed into the workflow with audit trails and human review, not bolted on afterward.

Further reading
AI & Operations

Eliminate, Automate, Too Late: How to Sort Support Work Before You Automate It

The EAT framework (Eliminate, Automate, Too Late) for deciding which customer support work to remove, which to automate with AI, and which to leave to people, plus how to classify customer pain at scale with an LLM.

AI & Operations

Building a Lead Signal Engine with LLM Classification

A working case study: designing, prototyping, and iterating an AI classification system that surfaces high-signal opportunities, including what broke and what finally made it useful.

Let’s find the first workflow.

Thirty minutes. Bring the process that eats the most time. You leave with a view on whether AI can help and what it would take.

Book a 30-minute call