How M2 freed coaches to scale instructional excellence

Dr. Pallavi Aggarwal was partnered with The Biome School in 2023 as part of a University of Missouri-St. Louis postdoctoral fellows program where she was introduced to the then–instructional coach–now Principal, Laura Myers. From their earliest observations, they realized that many teachers were struggling with basic Tier 1 instruction. Laura knew it wasn’t apathy that the teachers were experiencing; there were simply no systems in place to drive the kind of standardization around excellence they knew every teacher was capable of. 

As Laura recalls it, “The teachers who were well-prepared and had sharpened their lesson plans: they sparkled. It was the teachers who did not prepare, who would just fly by the seat of their pants – that’s where we noticed behavioral problems, teaching performance issues, and a correlation in lower scores out of those students.”

Pallavi also remembers those days. According to her, “Teachers weren’t required to review their teaching manuals, there were no PLCs for collaborative planning, and there was not a consistent instructional coaching model to guide growth.” 

99 observations, 44 videos, and one unsustainable path

Determined to shift the school’s trajectory, Pallavi and Laura threw themselves into data collection. In their first semester, they conducted 99 observations, meticulously documenting which instructional practices correlated with higher levels of student engagement. They presented their findings to leadership, making the first real case for systemic change.

In the second semester, they introduced GoReact’s video review platform to help teachers reflect on their practice and give leadership visibility into classroom instruction. But this tool came with baggage. After manually viewing and meticulously annotating video observations, Laura and Pallavi were craving some efficiency. Pallavi put it plainly,

“We recorded 44 videos and annotated every minute. It was a lot of work.”

In the same year, they piloted Swivl robots, which leveraged multiple microphones to capture classroom audio and provided speech analysis. The data was eye-opening: in most classrooms, teachers talked 98% of the time while students had opportunities to engage only 2% of the time. It was another data point confirming what they already knew – instruction needed to evolve.

The work paid off initially. By the end of that first year, they had developed three critical tools: a lesson template outlining how instruction should flow, two rubrics to assess instructional practices and student engagement, and a shared understanding that teacher capacity-building needed to be the school’s priority. 

Time to scale…and soar

Armed with their custom rubrics and building a culture of trust and partnership, Pallavi and Laura entered their second year ready to implement a more formal coaching model. But even with plans in place, the thought of manually observing, recording, annotating, analyzing, and providing personalized feedback to every single teacher each week loomed over them again like a dark cloud. They’d need to clone themselves just to maintain what they’d started.

That’s when Pallavi heard about M2 in a meeting with other mentors in her post-doctoral program at UMSL.

She was excited by the possibilities immediately: “M2 was a perfect fit in our trajectory because we were already conducting observations regularly, and fortunately, our teachers were amenable to trying new things.” But even if observations were not being standardized and teachers weren’t being served by the dream team that Laura and Pallavi became, M2 reinforced a very crucial principle for the teachers at Biome Charter: 

M2 observations are never meant to be punitive. They are designed to support growth. So the teachers approached M2 with curiosity, rather than resistance.

Using AI to bring out the humanness in their coaching

What immediately appealed to Pallavi and Laura was that M2 didn’t compete with their roles as coaches and mentors for their teachers. If anything, it became their trusted sidekick for every observation.

Here’s how their hybrid model works: M2 is set up in a classroom at the beginning of each lesson, collecting real-time data on the teachers’ skills: engagement, questioning strategies, and pacing. M2 is guided by the custom rubrics Pallavi developed to assess instructional practices so it knows what to look for and how to support each teacher during their class. After each lesson, M2 provides automatic scores, summaries, and actionable tips for improvement, while a human observer (usually Pallavi) adds context from their perspective of the class.

Teachers then review both sets of feedback and select their own instructional goals. “It’s not just us providing feedback. Teachers are also taking ownership when they view M2’s recommendations,” Pallavi explains. “From M2’s feedback, our teachers select two instructional goals for their next observation and then the cycle repeats. It’s so streamlined!”

The two leaders shared an anecdote that literally had them high-fiving as we spoke about it. One of their teachers, Dr. Washington, has become a model for how powerful this approach can be. Pallavi shared, “She’s actually using the tool in the best capacity possible. She carefully and thoughtfully reads through all of M2’s feedback and it helps her see the bigger picture.” 

Successful experiences like Dr. Washington’s are what make Pallavi and Laura’s partnership so strong and in sync: when they see a teacher take ownership of their own growth and development, it makes coaching and supporting that teacher even more rewarding.

Other teachers are becoming more vocal about their observation protocol too.

“They want to see their scores on engagement, questioning, and wonder when I bringing M2 back to their class.”

Dr. Pallavi Aggarwal

The next frontier: co-teaching with M2

Pallavi and Laura’s vision for M2 extends beyond observation and feedback. In Spring 2026, they plan to utilize M2’s Group Guide. With this feature, students can interact directly with M2 to ask questions, engage in small group discussions and receive feedback, and practice their skills, all while M2 provides instructional support and narrated directions.

They’re also exploring using M2’s customized assignment feature, which lets students verbally share their learning based on prompts that M2 designed from the lesson. “It will be really interesting to assess students’ depth of knowledge,” Pallavi muses, thinking about the various ways M2’s software [MirrorTalk.ai] will be able to analyze student reflections.

What M2 was designed for

When asked what guidance they’d give to other schools, researchers, or coaches considering M2, both Pallavi and Laura immediately emphasized the importance of relationships, which is also fundamental to the M2 experience. 

“The most important thing is to invest time in building relationships with each and every teacher, reminding them that M2 is not evaluative; it’s for the betterment of their teaching and students’ learning.”

Laura Myers

The cultural groundwork matters. Trust matters. And when those foundations are solid, technology like M2 can transform what’s possible–not by replacing teachers or coaches, but by giving them the time, data, and support to do their most important work: helping every teacher grow.