Memoir

Inspiration

Memoir started as a very personal problem: my memory has been betraying me for as long as I can remember.

At school, it showed up as struggling to retain what we were learning. I have ADHD, and I always felt like I needed to work twice as hard just to remember what others seemed to retain naturally. I tried every technique: rewriting notes, repeating information out loud, re-reading textbooks. Sometimes it helped a bit, but it never really fixed the core issue.

As I grew up, the problem moved from grades to something that felt much more painful: my relationships.

I started forgetting things that, to me, define being a good friend or partner:

• parents’ names
• where someone went to college
• the small details they shared in moments of vulnerability
• important dates and milestones

I’d find myself in conversations feeling a mix of anxiety and shame because I couldn’t recall things that clearly mattered to the other person. I tried to “hack” this by repeating information to myself, or writing things down in my notes app. At one point I seriously considered building a full-on Excel CRM for my friends. But none of it was sustainable. CRMs are painful to maintain, and they don’t proactively help you remember what matters.

The most painful moment came about a year ago: a very close friend of mine lost one of her twins. I wanted so badly to be there for her in a thoughtful, personal way; to remember her baby’s name, to remember the week this happened so I could check in, send a message, honor that loss. And I found myself struggling to recall those details. That felt… wrong. Not because I didn’t care, but because my memory simply failed me in a moment where empathy should have been automatic.

That’s when the “Excel friend tracker” stopped being a joke and started feeling like a necessity. But I also knew there had to be a better, more human way.

As I talked to more people, I realized this wasn’t just “my” problem:

• One friend writes every experience in his phone notes because he’s afraid of forgetting his life.
• Another writes down her dreams every morning.
• Many people try to journal, but struggle to keep the habit.
• Others feel overwhelmed by scattered memories: photos in one place, notes in another, voice memos somewhere else.

Everyone was building their own fragile personal systems to remember what matters — and none of them really worked.

That’s how the idea for Memoir was born.

It started as “a CRM for the heart”, a way to remember details about friends and loved ones. Then it expanded: What if we had an app that could become a digital extension of our mind?

An app where you can capture memories in whatever way feels natural (voice, text, photo), that then organizes, labels, and connects them, and proactively brings back the people, moments, and feelings that matter.

Not just data storage. A memory companion.

What it does

Memoir is an AI memory companion that helps you capture, organize, and recall what matters in your life.

In our hackathon MVP, you can:

• Capture moments effortlessly:

  Speak a quick voice note, type a thought, or paste a snippet. Memoir transcribes voice automatically and cleans it up.

• Let AI do the organizing:

  Each “Moment” is processed by AI to extract: 1. people (e.g., “Sofia”, “Mom”) 2. places (e.g., “Lisbon”, “Chicago”) 3. time references (e.g., “last summer”, “two years ago”) 4. themes and emotions (e.g., “grief”, “gratitude”, “celebration”) Memoir then tags the memory so you don’t have to manually organize everything.

• See your memories in context

  Moments are shown in a simple, clean feed with their tags and emotional tone. You can filter by people, emotion, or time (“show me moments with X”, “show me happy moments”, etc. in future versions).

• Get recall prompts when you need them

  Memoir can surface simple recall cards like: 1. “You mentioned your friend’s product launch last week — do you want to check in?” 2. “Around this time last year, you were going through a big change. Want to revisit that?”

The long-term vision is a Memory Graph that connects people, places, and feelings across time; but for the hackathon, we focused on shipping a small, lovable slice: capture → organize → recall.

How we built it

We built Memoir using Lovable’s no-code/AI platform to move fast while still shipping something real.

At a high level:

• Frontend in Lovable

  We designed a minimal web interface with three core screens: 1. Capture: a page where users can add a new Moment via text or voice. 2. Memories Feed: a scrollable list of Moments with tags and emotions. 3. Recall: a simple “suggested memories” view powered by AI. Lovable generated the base React components, and we iterated on copy, layout, and UX to keep it as frictionless as possible.

• Backend & data model:

  We created a simple data model around User, Moment, Tag, and Person. Each Moment stores the raw input, the cleaned transcript, detected entities (people, places, dates), and emotion scores. Data is persisted in a database managed through Lovable so the app feels like a real product, not just a mock.

• AI layer

  We used an LLM to: 1. Transcribe voice notes. 2. Extract entities (names, locations, dates). 3. Infer emotional tone and themes. 4. Generate human-friendly recall prompts. This AI layer sits behind a simple API endpoint that the Lovable frontend can call.

• Iteration and testing

  We seeded Memoir with our own real moments to make the demo feel alive. We then iterated on the prompts and UI to make outputs feel more emotional and less robotic.

In short: we used Lovable to do the heavy lifting on scaffolding and wiring, and focused our time on product experience and AI behavior, not boilerplate code.

Challenges we ran into

We ran into both technical and human challenges.

1. Where and how to store such sensitive data If Memoir works, people will store their most intimate memories in it. That raises big questions: > How do we store data securely? > How do we design for privacy from day one? > How should we think about encryption, access, and “right to forget”?

For the hackathon, we kept things simple but are already designing for a future where data is encrypted and controlled fully by the user.

1. Defining the first problem to solve “Memory” is huge. We could focus on: > relationships > life journaling (i.e., experiences) > dreams > creative ideas > health & mood tracking, etc

Trying to solve everything at once is a recipe for confusion. We had to narrow it down to “help me remember what matters in my everyday life” and show that clearly in the MVP.

1. Habit formation and logging behavior Even with AI, people still need to log new information. Questions we wrestled with:

   How do we make capturing a moment feel as easy as sending a voice note to a friend? How do we reduce friction so logging once a day feels natural, not like “doing homework”?

2. The past vs. the present It’s one thing to log new moments going forward. It’s another to ask: “What about my entire past?”

Do we ask people to backfill? Connect photos? Import chats? We explored the idea of gently prompting users to recall key chapters of their life over time, instead of expecting them to dump everything at once.

1. Building the right team Memoir sits at the intersection of AI, psychology, behavior design, and product. Finding the right people who care deeply about both tech and human emotion is an ongoing challenge — but the hackathon helped us bring the first version of that team together.

Accomplishments that we're proud of

• Turning a personal insecurity into a product vision. This started as “I’m bad at remembering things and I’m ashamed of it.” It’s now a concrete product that others can use and react to.

• Real validation from real people. As we started sharing the idea, people leaned in. Friends and classmates said things like:

  “This would literally change my life.” “I journal all the time but still can’t find anything later.” “I’m terrified of forgetting important moments.”

• A working prototype that feels like the future. Using Lovable, we shipped a real, interactive demo where users can actually record moments and see AI-organized memories — not just slides.

• Clarity that this is bigger than a hackathon. The more we worked on Memoir, the more convinced we became that this is not just a weekend project. It’s a problem that touches almost everyone.

What we learned

• “What matters” is deeply personal. For some, it’s friends and relationships. For others, it’s creative ideas, spiritual experiences, or health milestones. One-size-fits-all structure doesn’t work. Memoir needs to be flexible and respectful.

• Understanding users at a deep level is non-negotiable. We can’t just guess what people want to remember. We need ongoing qualitative interviews and behavioral data to see what sticks and what doesn’t.

• Memory is not just data — it’s emotion. A basic notes app can store information. The real magic is in helping people feel connected to their past selves, their relationships, and their growth.

• Building trust is as important as building features. If people don’t trust Memoir with their memories, they’ll never use it, no matter how “smart” it is. Privacy, clarity, and honesty have to be core features.

What's next for Memoir

After BoothHacks, we want to take Memoir to the streets:

• Deep user research at scale

  Run surveys and interviews to understand what people actually want to remember. Collect both quantitative and qualitative data to see patterns across users.

• Private beta with real users

  Start with a small group of early adopters (e.g., classmates, journaling enthusiasts, busy professionals). Watch how they use Memoir, where they drop off, and what moments feel most meaningful.

• Refining the product around real behaviors

  Use feedback to adjust the capture flows, tags, and recall prompts. Experiment with different nudges: daily check-ins, weekly recaps, anniversary reminders, etc.

• Hardening the technical architecture

  Design a secure, privacy-first foundation for storing memories. Move toward hosting models and data in a controlled environment with a closed loop for user information. Explore on-device or end-to-end encrypted approaches so users can feel fully safe.

Most importantly, we’ll keep using tools like Lovable to quickly iterate on something tangible. Having a real, clickable version of Memoir that people can try unlocks better conversations, better feedback, and faster learning.

Memoir’s long-term mission is simple but ambitious: Help humanity remember what truly matters. This hackathon is our first concrete step.

Built With

• lovable
• openai