Not slideware — measurable throughput, cost, and revenue.
I'm Chris Tassoulas — a Lean Six Sigma Black Belt, SAFe Lean Portfolio Manager, and former COO who rebuilds how operations run and wires generative & agentic AI into the core of them. I've scaled a startup as an operator and I now help run the performance engine of an $84B-asset bank network. Whatever hat the mission needs — operations, portfolio, or product — I plug in and ship results.
The same core — measure, model, automate, govern — pointed at whichever role the organization needs filled. Each is backed by real credentials and shipped work.
Re-engineer end-to-end operations and embed AI + agentic automation to cut cycle time, error, and cost — then prove it in the numbers.
Govern the portfolio of AI initiatives — fund the right bets, sequence delivery, connect Lean budgets to outcomes, and keep enterprise-scale programs moving.
Translate messy business needs into AI products and agentic workflows — backlog, BDD stories, RAG/agent design, and shipped features people actually use.
Three working prototypes tied to real operations problems I own. Move the controls; the logic runs live in your browser.
Retail-bank incentive plans quietly leak money: payouts drift from real attainment, and outliers hide across dozens of centers. Today that's caught by hand, late. This flags it the moment the numbers move.
I understand incentive-comp mechanics and the statistical anomaly detection that ML monitoring is built on — the exact intersection of my current bank role and applied AI.
Staffing and branch hours are usually set by tradition, not demand. This models a center's hourly demand and recommends the staffing — and the open/close hours — that hit service targets at the lowest labor cost.
The workforce-management, staffing-model, and banking-center-hours analysis I run today, expressed as an optimization engine instead of a spreadsheet.
Frontier-grade agentic AI isn't one prompt — it's a governed orchestra: a supervisor decomposing the task, specialist agents retrieving, analyzing, and drafting in parallel, then guardrails, evals, and a human gate before anything executes. This runs that architecture on enterprise & public-sector workloads.
I architect production multi-agent systems the way frontier labs and enterprise AI teams do — supervisor/worker orchestration, RAG with citations, guardrails & evals, full observability (tokens · cost · latency), and human-in-the-loop governance.
A portfolio always has more AI ideas than budget. This ranks candidate initiatives by WSJF — weighted shortest job first — and funds down the list until the Lean budget is spent. Move the budget and watch the cut line move.
I can run a Lean portfolio: score value against cost, sequence the funded set, and defend the cut line. SAFe LPM applied to a real AI roadmap.
From the floor of a scaling startup to the analytics core of a national bank — the same operator discipline, quantified.
I own workforce-management modeling, the staffing model, and banking-center-hours analysis across the network; direct goal modeling and allocation for the Consumer Bank; and administer the incentive-compensation program — plan documents, goal setting, payout accuracy, and reforecasting to keep the line of business inside plan. I turn transaction and staffing data into the reporting leadership uses to make strategic calls on hours, coverage, and incentive spend.
Owned P&L, operations, and digital transformation. Re-imagined core workflows with customer-journey mapping, redesigned KPIs and SLAs, ran the SAFe/Agile SDLC for critical digital initiatives, and led legacy-system migration plus Avalara/Zonos integration for international expansion. Built the SOPs, dashboards, and cadence that let the company scale.
A repeatable loop that fuses Lean Six Sigma rigor, SAFe portfolio governance, and applied AI — so automation lands on measured problems and stays accountable.
Define the problem in numbers first. Baseline cycle time, cost, error, and value with the same discipline as a Six Sigma DMAIC define/measure phase.
Model the process and the decision. Where does judgment repeat? That's where an agent, a model, or an optimizer earns its place.
Build it — RAG retrieval, agentic workflows, optimization — with a human-in-the-loop gate on anything that touches money or risk.
Fund it, sequence it, and prove it at portfolio scale. Lean budgets, guardrails, and the metric moving in the right direction.
Operational rigor, portfolio governance, and applied AI — the credentials behind the work.
I'm exploring senior roles in AI operations, transformation, and product — and select advisory work. If you're wiring AI into how your business actually runs, let's talk.