AI in practice

Useful experimentation. Responsible adoption.

I start with a real user or workflow problem, then balance the opportunity with evidence quality, usability, governance, cost, and deployment controls.

Try the Role-Fit Agent Ask Wayland AI

My build thesis
“I don’t just advise organizations about AI adoption. I build and experiment with AI products myself.”

Enterprise AI adoption

Business problems first. Scope and controls made explicit.

At Standard BioTools, I delivered an executive insights agent in business use, built workflow agents, and evaluated the conditions needed for broader analytics adoption.

01Working · In business use

Executive Insights AI Agent

I delivered conversational access to financial information for the CFO, CEO, and regional leaders. Business users are using the agent.

How it works

The agent uses existing Power BI semantic models through an analytical workflow implemented with Copilot Studio and Power Automate.

Parameterized DAX queries, structured JSON responses, access restrictions, and regression tests are functional.

COGS, OpEx, pipeline, and other expansion domains remain future scope.

Business use is confirmed; usage by each executive role, rollout scale, and measured time savings are not established.

02Built · Adoption unmeasured

D365 & Power BI Workflow Agents

I built agents for D365 backend-query support and Power BI ticket intake, addressing query-writing effort and duplicate support-request friction.

The builds are confirmed. Production scale, adoption, and realized time savings have not been measured.

03Evaluation · Hands-on trial

Fabric Implementation Evaluation

I evaluated Microsoft Fabric’s integration, release controls, Power BI Copilot, and licensing options, and created a capacity-testing roadmap with defined success criteria.

What I evaluated

Integration and delivery: I evaluated new data-flow patterns, Dataflow Gen2, XMLA endpoints, data-lake integration, deployment pipelines, and Git-based version control.

Power BI Copilot: I tested the capability inside Power BI/Fabric, distinct from general-purpose Microsoft Copilot. I used F2 as a budget-conscious paid entry point because the trial did not include the Power BI Copilot experience I needed to test.

Workload and licensing: I assessed real workload requirements and capacity behavior to inform the appropriate license and production capacity. F2 was the test entry point, not an assumed production target.

Hands-on trial: I applied ALM Toolkit, Tabular Editor, and selective SSMS partition refreshes to improve efficiency. No measured time-saving result is established for this trial.

Evaluation and development work; no completed enterprise migration or production rollout is claimed.

What the work demonstrates

InitiativeProblem → solutionAdaptabilityResponsible adoptionProduct thinking

Independent AI product · 2026

A working product, built with AI assistance.

I built and published waylandzeng.com using AI-assisted development, structuring career evidence and case studies for recruiters and designing grounded career Q&A and role-fit features.

What the project demonstrates
Define a user need, organize trustworthy information, build a useful experience, and improve it through iteration.

01 / Problem

A static résumé cannot answer the next question.

Traditional résumés and LinkedIn profiles give recruiters limited time and limited ways to explore the depth of my transformation experience.

02 / Product decision

Ground the experience before making it conversational.

I structured verified career evidence as the source of truth, then defined recruiter workflows, career-only scope, refusal paths, and clear distinctions between verified and unsupported claims.

03 / Build approach

Move from an ambiguous idea to working releases.

I used AI-assisted development to design, build, publish, and iterate the platform while retaining ownership of the product requirements, evidence boundaries, user experience, safeguards, and release decisions.

04 / What shipped

Two bounded recruiter experiences.

Ask Wayland AI answers approved career questions. The Role-Fit Agent compares a job description with the verified record and separates direct matches, transferable evidence, and requirements not evidenced in that record.

What the build reinforced

Useful AI adoption starts with a defined user problem and trustworthy context. The model matters, but so do evidence design, boundaries, usability, privacy, abuse controls, cost, and the operating model around it.

Role-Fit Analysis Agent

Start with what the role needs to accomplish.

Compare the role’s business mandate and responsibilities with my documented experience, then explore transferable strengths and questions for an interview.

Live · Career questions only

Ask Wayland AI

Have a question about my experience?

Explore my professional experience through an AI assistant grounded exclusively in approved career information.

For approved career questions. The assistant may make mistakes; verify important details against the resume or contact me directly. Questions are sent to OpenAI to generate answers; this site does not save question text in its database.

Ask Wayland AIVerified career information only

Live

What would you like to understand about Wayland’s leadership, transformation experience, working style, or professional interests?

Discuss an opportunity

Interested in discussing practical AI adoption or the thinking behind this product?

Let’s talk about the business decisions, team, and reporting challenges your role needs to address.

Email Wayland Download 1-page Recruiter Brief Connect on LinkedIn Download resume PDF Download Resume Word