Order.co Innovation & Technology Culture

What tools support your day-to-day work?

Across the company, teams leverage AI tools like ChatGPT and Claude to accelerate research, content creation, problem-solving and workflow design. Recently, Convey has become one of the most impactful tools in our day-to-day work, empowering technical and non-technical employees to create AI agents that can automate processes that previously required significant manual effort. The biggest benefit on top of productivity is giving people more time to focus on strategic, high-value work rather than repetitive tasks.

How does your team experiment?

We take a hands-on approach to experimentation by encouraging every team member to explore the automation and AI tools available to them. Rather than limiting innovation to a single team, we empower individuals to identify inefficiencies, test solutions and share what works. We've already seen significant results, from eliminating time-consuming manual processes to creating entirely new capabilities. One example is our design team building a custom Claude skill that can instantly apply our brand standards to internal content and external assets, saving time while maintaining consistency.

How does your company adapt to change?

One of the reasons we're able to adapt quickly is that change often starts with grassroots adoption. We equip teams with the right tools, encourage experimentation and create opportunities to share successes across the organization. A recent example was our transition to a new HRIS platform that consolidated multiple people systems into a single experience. The platform's built-in AI capabilities allow employees to quickly access information, complete common tasks and find answers independently, improving efficiency while creating a better employee experience. By pairing new technology with employee-led adoption, we've been able to drive meaningful change across functions.

Can you share an example of how you’re using AI in your day-to-day work and what problem it helps you solve?

My favorite example is our weekly team recap. The hardest problem on a data team isn’t writing code, it’s shared context — what shipped, what’s blocked and why. Historically, that has been scattered across Jira, GitHub, Slack and a dozen pipelines. Assembling that picture used to eat half a day and was still incomplete. Now an agent pipeline does the discovery. It pulls delivery status from Jira, activity from GitHub, signals from Slack and our coding-session metrics via MCP; cross-referencing all and then drafting an evidence-linked recap. Half a day of compiling became minutes of editing and every Monday the whole team starts from the same picture of reality. The pattern we profess is: agents automate discovery so humans start from synthesis instead of scavenging. We now maintain a shared toolkit of 50-plus agent skills, like triaging an ETL failure directly from a Slack alert, querying Snowflake through Cortex or authoring a PR automatically to our convention standards. New teammates can bootstrap all of this setup in two commands. The ultimate payoff is decreased time to delivery on everything we do: Work starts at “root cause found, fix proposed,” not “let me go look.”

How has your company supported you in experimenting with AI or building new AI-powered solutions?

The support has been comprehensive: from budget to trust and celebration. Everyone who wants them gets paid seats for frontier tools like Claude Code, Cursor or model API access. Usage is encouraged rather than rationed; AI-assisted work is expected to be the default operating model, not a nice to have. The deeper support is trust in production. Our team ships AI as core infrastructure: semantic search over a catalog of a million-plus products, LLM-powered extraction pipelines and AI built dashboarding. When your company lets AI touch production systems, with the guardrails to earn that trust, experimentation stops being confined to a sandbox activity. The experiments compound, too. We point agents at most open questions, not just engineering chores. One engineer went from a raw question to a first class analysis with every claim evidence backed and an executive-ready product brief in two weeks, work historically took months. Results like that drive roadmap decisions and the company gives internal awards for multiplier-style AI work. The result is culture rather than policy — nobody asks permission to try AI on a problem. They ask whether anyone has written a skill for it yet.

How has working with AI changed the way you think about your own career or future growth?

I started my career in finance 25 years ago with no technical background. I learned SQL on the job at TripAdvisor and for years my Python was, charitably, rough. Today, I ship hardened production systems — pipelines, agent harnesses, guardrail automations and its routine. Not because I finally became a great programmer, but because AI collapsed the distance between what I can judge and what I can build. Waiting on execution stopped being my bottleneck. That experience rewired how I think about growth. What we build is becoming a commodity. The same pipeline can come from a teammate, an agent or a fleet of agents running overnight. The durable skill is strategic thinking. Choosing the right problem, framing it precisely enough that any executor can nail it and designing the verification so you can trust work you didn’t do line by line. Those are the muscles I now train deliberately and the ones I hire for. The arc I lived, strategy brain first, technical skills late, used to be a handicap in engineering — AI flipped that. I truly believe that the future belongs to people who know exactly what to build and can now, finally, just build it.