Eva brings data-driven quality control to school photography

In volume school photography, image quality is everything — yet most studios still rely on manual review processes that struggle to keep pace with thousands of images arriving every day. That challenge is exactly what 36Pix CEO Robert Ste-Marie set out to solve with Eva, a new AI-powered platform designed to evaluate photographer performance in near real time.

Ste-Marie, whose Quebec-based company photographs more than 600,000 students annually across Canada, knows firsthand the pressures facing high-volume studios. During peak season, his operation processes between 10,000 and 25,000 images per day from a team of more than 125 photographers. With annual photographer turnover running as high as 50–55 percent, training and quality control become mission-critical.

“In school photography, the photographer is the product,” Ste-Marie explains. “If the image quality isn’t good, sales suffer. And if things really go wrong, you can lose contracts.”

Historically, quality control has relied on staff manually reviewing folders of incoming images — a process that’s time-consuming, inconsistent, and prone to missed problems. A photographer could be producing subpar work for days before issues are identified, potentially impacting multiple school accounts before corrective action is taken.

That reality led Ste-Marie and his team to build Eva — short for “evaluation” — an AI-driven platform that automatically analyzes every image as it enters a studio’s workflow. Instead of sampling a few folders or reacting to parent complaints, studios can now monitor image quality continuously and objectively.

Eva evaluates six key quality factors: exposure, glasses glare, smiles, closed eyes, blurriness, and facial shine. Each image receives a score, and studios can assign custom weightings depending on the job type. For example, smile detection might be weighted lower for high-school portraits and higher for elementary schools.

“AI doesn’t care who you are,” Ste-Marie says. “It evaluates everyone the same way. That removes bias and gives you real data to work with.”

The system operates in near real time, typically within 24 hours of capture. That means studio managers can identify trends quickly and intervene before problems escalate. If a photographer is consistently producing blurry images or struggling with lighting, managers can provide targeted coaching or training immediately — not weeks later at the end of the season.

One of the biggest surprises for Ste-Marie has been smile detection. After reviewing a full season of data, he discovered wide variability among photographers — and a higher-than-expected number of students not smiling, even in elementary schools.

“When you see that some photographers are getting 10 percent non-smile rates and others are at 35 or 40 percent, that tells you something,” he says. “That’s a training opportunity.”

Eva also gives photographers access to their own performance dashboards, comparing their results to company averages. That transparency encourages self-reflection and healthy competition while eliminating the perception of being unfairly singled out.

“It becomes about the data,” Ste-Marie says. “What can I do better?”

The platform also provides studios with powerful tools for account management. If a school raises concerns about image quality, managers can instantly review that job’s data, assess photographer performance, and respond with facts — not guesswork.

From a business perspective, Eva is offered on a per-image pricing model, reflecting the cloud computing and AI processing required for each evaluation. Onboarding typically involves linking photographers to their camera hardware IDs or file-naming conventions, allowing images to be automatically attributed to the correct shooter.