In this issue: 

Read time: 7 minutes

When AI makes it possible to build almost anything, how do you decide what’s worth building? And how do you know it will last, and can’t easily be copied?

Stuart Parmenter, former CTO of One Medical, an early Firefox engineer, and an advisor to Reach, has written at length on this question. His answer: as code gets cheaper, the advantage shifts elsewhere: judgment, trust, customer obsession, relationships, and how quickly a company learns.

It’s a playbook that Mara Steiu has instinctively followed throughout her life. Early on, she already knew what she was: a builder. By the time she graduated high school in Romania, she had already started two education organizations.

After college, customer obsession guided her to her next entrepreneurial venture: Journify Learning, an AI assistant for special education that automates tedious paperwork, simplifies compliance efforts, connects data so that teachers can give students the support they need.

It all started with over 100 conversations with teachers, principals, and superintendents and one question: Where do you feel the greatest pain points?

We sat down with Mara to talk about how student builders can learn firsthand from users and bake those learnings into a product that users trust and return to.

Reach Capital: How has AI changed the way you think about what’s worth building?

Mara Steiu: It’s so easy to build almost anything instantly. But I actually think that can make things riskier for a founder. When resources feel unlimited and anything is cheap or almost free to build, you don’t have that same pressure to identify what actually drives value and what a customer really needs.

If I were starting today, I would still place that pressure on myself. Time is very precious and very limited. I would act with a resource-constrained mindset because that pushes you to prioritize ruthlessly and be extremely customer obsessed and disciplined.

It matters as you grow, too. If a customer has a product that does 100 things and they only use two of them, they’re going to ask, “Does it make sense for me to pay for this? I’m only using 10% of it.”

Just because it’s cheap and you can build more things doesn’t mean building more is better. You have to understand the core value and double down on that.

How do you decide what to prioritize?

A lot of it goes back to observing the customer.

For us, we observe teachers using the product every week through our advisory board. We don’t say anything.

If they stop and say, “This doesn’t make sense,” that’s when we listen in and ask, “Okay, why doesn’t this make sense?”

Of course, we also interview users and track product usage. But observation in the flow of work is extremely important. You can see things someone might not think to tell you.

What did you observe from growing up with teachers yourself?

My mom is still a public school teacher. I see her go to school in the morning, come back, prepare lesson plans, grade materials and keep working until eight or nine at night. It can basically be a 12-hour job. It’s such an intense job.

That’s what I try to keep in mind: teachers are extremely busy. They have full days of work from morning to night, and they have zero time for new tools. Anything they adopt has to substantially make their lives better.

I’ve also noticed with my grandmother, who was a public school teacher, and sometimes my mom, that it’s not necessarily a fear of technology. It’s more the adoption curve: Do I have time to learn this?

So ease of use is a big design principle for us: How do we make every new functionality as intuitive as possible?  How do we make it smooth to get started? And how do we observe how teachers use the product so that if we see blockers, we can fix them quickly?

Journify is your third startup. What did you build before then?

My first startup was in the financial literacy space. We built a solution for schools in Romania, and that was my first experience with edtech and entrepreneurship.

I loved it so much that before moving to the U.S. for university, I started another initiative: a nonprofit startup accelerator for high school students in Romania. As a high school founder, I realized the public education system offered little support [as an entrepreneur]. I wanted to make sure other high school students had access to the resources they needed to build their own companies. We ended up helping more than 500 students build their own companies. It was an amazing experience.

The biggest lesson was simply that I could build something and bring ideas to life.

What did you take from your time at big tech?

I knew I wanted to go back into building in K-12, but I knew I needed more technical skills. I wanted to understand what it takes to build at scale.

At Twitter, I learned a lot about culture. It felt a lot like a family, which I didn’t expect. Everyone was having fun working together. I remember thinking, Whenever I build my own company, I want to bring this type of culture and approach to work with me.

At Microsoft, I was a product manager for Azure SQL, which was heavily technical but also extremely customer obsessed.

You might think of Microsoft as a big established company and assume it isn’t entrepreneurial, but this team really was. Because Azure SQL is used all over the world, I could see the impact of building something and getting it into customers’ hands. Customer obsession was a huge lesson for me there.

And I also met my co-founder, Sagar Manchanda, at Microsoft. So that corporate route ended up being extremely beneficial for me as a founder.

How did you get the idea for Journify?

I was doing an MBA and an MA in education at Stanford knowing I wanted to build another company in K-12 public schools. My approach became: Where is the need the biggest?

So I started interviewing teachers, principals and superintendents just to understand their daily lives. I went into schools whenever I was allowed, observed their work, and mainly just did interviews. I think I spoke with around 100 people asking: What are the biggest challenges in your daily life?

Special education was the most common theme I heard, even though I wasn’t specifically interviewing special education teachers. These were general education teachers, principals and superintendents.

As I started looking deeper, I saw so many challenges happening in parallel. There were severe staffing shortages. Special education teachers were spending a huge amount of their time on paperwork and compliance. And there were major gaps in student outcomes.

Special education can look like a niche market, and I think that’s one reason a lot of edtech companies have avoided it, especially because of the compliance intricacies. But I don’t think niche markets are something founders should necessarily be afraid of.

How did you get educators and district leaders to talk with you while you were still a student?

I think being a student can be very powerful. When you’re a student, I’ve found that people can be much more open to talking with you and sharing what their challenges are and what they really think about the space. 

At the same time, you should think about what they’re getting in return. If you’re doing research, you should be willing to share your findings with them so they’re learning from you as well. Once you have prototypes, you can give them access for free so they benefit too.

Go in with honesty and curiosity. People are actually quite open to helping.

What did you learn from those early conversations, and how did that shape the first version of Journify?

From the very beginning, we co-built the product with two districts, one in California and one in Washington.

We worked with a few teachers in each district from scratch. They knew we were very early. They used the product for free and gave us a lot of feedback about what worked and what didn’t.

That helped us understand the problem much more deeply. We focused on connecting data across those support providers, helping teams monitor student progress, creating personalized instructional supports and helping school and district leaders with compliance.

Co-building the product from scratch with one or two customers was extremely important. We weren’t building something and coming back once it was finished. They were helping us understand what worked and what didn’t as we went.

What do you look for when you’re evaluating someone you might hire?

My co-founder, Sagar, has done a lot of our engineering hiring, but one thing we look at is what someone has actually done in the past and the kinds of projects they’ve worked on.

If we see an engineer who has worked in K-12, that’s a strong plus. It’s not required, but it can show that they really care about the mission in addition to having strong technical skills.

On the technical evaluation side, one thing we’ve tried is giving someone a task that may actually be impossible to solve, even with AI.

They might not be able to solve it. But we want to see if they can come back and say, “I couldn’t figure it out, but this is how I thought through it. These were the variables I looked at.”

And I don’t think that only applies to engineering. You could do something similar for sales or other roles. How someone approaches that problem end to end is really important.

A year ago, our partners Jennifer Carolan and Wayee Chu shared what they look for in exceptional interns.

This summer, Talia Buckhouse flips the perspective.

Instead of offering another list of interview tips or resume advice, she shares something more revealing: how she approached the experience itself.

The systems she built before anyone asked. The habits that made feedback easier to absorb. Why sending one-pagers changed the quality of her conversations. And why the most important thing she brought to Reach wasn't something she could fit in a backpack.

→ Add your builder profile to be matched with founders across the Reach Portfolio

→ AI Research Scientist Intern at GenMD (Month-to-month, Palo Alto)

→ Generalist Intern at Messa (Full-time, Remote)

→ Software Engineer at Desmos (Remote)

→ Software Engineer at Scout (New York)

→ Software Engineer I at Handshake AI (San Francisco)

Happy back to school season,

Valentina and the Reach team

Reach Capital invests in early-stage founders redefining how we learn, live, and work. Our portfolio of 130+ startups includes tools you might’ve used in school (ClassDojo, Desmos, Brilliant) and next-gen, AI-native disruptors shaping how future generations build, work, and thrive (like Replit and GPTZero). 

These teams are constantly on the lookout for talented builders. Share what you’re studying, building, or exploring, and we’ll intro you to projects, people, and paid opportunities for you to build upon.