5

If you apply any automation to a broken workflow, you aren't achieving efficiency, you are simply making mistakes faster.

According to Jennifer Chun, CPA on the finance team at 1Password, many organizations are currently putting band-aids to fix fundamental workflow issues. Jennifer mentions how before even taking a look at automation or integrating AI, we should take a step back and decide, “Should this workflow even exist in its current state?” In the context of today’s talent, tooling, and capabilities, we might be able to build up a more solid foundation than before.

In this episode of The Optimization Era, Jennifer pulls back the curtain on how finance leaders can transition from anxiety-driven automation to a strategy rooted in analytical thinking and documented truth.

Who is Jennifer?

Jennifer Chun is a Canadian CPA and a key member of the revenue accounting team at 1Password. Beyond her core work in revenue recognition and monthly closes, she has spent the last six months serving on 1Password’s AI and Finance Committee, helping data teams become fluent in AI enablement.

What 1Password does: 1Password helps enterprises manage human, agent, and machine identities in a world where AI accelerates both productivity and risk. Their recently launched platform, 1Password® Unified Access, surfaces AI tools and exposed credentials across endpoints, and vaults them into one secure system, with governance applied to high-risk and shared accounts.

Diagnosing the "Jell-O" Foundation: Why Workflows Need a Redesign

The biggest barrier to AI ROI isn't the technology, it's the invisible work hiding inside outdated spreadsheets and judgment-based conversations. Jennifer argues that many companies are still running on workflows designed years ago for a much smaller scale.

"If you apply a new Band-Aid approach to it... you are just making mistakes faster. On the surface, it can look like progress... but the underlying process could still be broken unless you look at it first.” – Jennifer Chun, CPA

In a conservative field like finance, where reputation is the primary currency, the risk of invisible errors is high. If an AI tool misses the nuances of a materiality discussion or a back-and-forth context living in an accountant's head, the output might be fast, but it will be wrong.

Jennifer’s solution is a fundamental perspective shift: stop asking what can be automated for ROI, and start asking why this workflow exists in the first place.

Surfacing the Truth through Documented Reality

Before you can improve a process, you have to see it. Jennifer is a firm believer in documenting the current state before asking how to fix it.

She shared a humbling realization after using Scribe to document her own work: the tool titled her process "copy and pasting". It was a stark reminder that even high-level roles are often bogged down by non-value-added manual tasks.

Jennifer provided the following frameworks for how she is approaching AI and automation to see where it can truly have the most impact:

Make the Work Physical: You must surface the judgment calls and exceptions that currently only live in people's heads.
The Junior Accountant Lens: Treat AI tools like an intern or a junior staffer who is completely new to the company. If you don't define what good quality looks like or specify the audience (e.g., a concise CFO vs. a broader accounting group), the AI will miss the point.
Communication is the New Technical Skill: Jennifer has observed that the strongest communicators get the best results from AI. They are better at articulating expectations and tailoring context, which drastically reduces the time wasted on generic, useless output.

Pro-tip: Use Scribe Optimize to identify those copy and paste black holes in real-time. By mining process data as it happens, you can distinguish between idealized workflows and the actual reality on the ground, identifying which repetitive tasks are truly ripe for automation.

Killing the CTRL+F Nightmare: A Success Story

Not every AI project is a winner. Jennifer spent a week trying to build a custom GPT to fill out revenue packets, only to realize that the manual effort of gathering the data was the real bottleneck, not the typing itself.

However, her team saw a massive win by automating Master Service Agreement (MSA) reviews.

Historically, junior accountants had to manually Control+F through legal contracts to find termination clauses or pricing caps. By building a custom GPT that provided a structured summary in seconds, the team didn't just save time; they freed up mental capacity for higher-value analysis.

"One of [the junior accountants] told me he cannot imagine not having this... he would never consider going back to Control+F". – Jennifer Chun

Her advice? Start with the most repetitive or annoying tasks in your day. There’s a high chance you can build a simple automation to start your journey to becoming an AI-first company and see some real ROI gains. Start small, then build up once you have a good foundation of testing!

Why the Human in the Loop is the Ultimate Safe Guard

As AI agents become a reality, Jennifer remains an advocate for analytical thinking (or "healthy skepticism"). AI can be a "Hype Man," telling you every idea is excellent and overconfidently buttering you up.

In the Optimization Era, someone must always put their name to the work.

"You need someone to blame when things go wrong," Jennifer notes. "Who are you gonna blame? Your Gemini bot?"

Beyond accountability, she believes the social interaction of collaboration is a biological necessity for humans; one that technology should augment, not replace.

Starting from Zero AI? Check out Jennifer’s Advice!

If you are starting from zero today, Jennifer recommends looking at people, not just tools. Success in the Optimization Era is a matter of change management:

Identify the Hand-Raisers: Find the people in your organization who are already intrinsically motivated and curious about AI.
Unfollow the Hype: If an algorithm or influencer is feeding you fear and anxiety, unfollow them. Focus on practical use cases, not noise.
Audit the Copy and Paste: Look for those non-value-added manual processes that everyone hates. Document them, then ask: "Does this workflow even need to exist?"

When you understand your workflows at a microscopic level, you stop building on Jell-O and start building a future that is sustainable, secure, and, above all, human-led.

Catch the full session below! 👇