Our Approach

01 /

To minimize burn rate, we bootstrapped and kept the team small: just myself handling Design and Product, while my co-founder, an ex-YC alum with a background in Finance, managed Engineering and Business.

Our Approach

01 /

To minimize burn rate, we bootstrapped and kept the team small: just myself handling Design and Product, while my co-founder, an ex-YC alum with a background in Finance, managed Engineering and Business.

Our Approach

01 /

To minimize burn rate, we bootstrapped and kept the team small: just myself handling Design and Product, while my co-founder, an ex-YC alum with a background in Finance, managed Engineering and Business.

Our Approach

01 /

To minimize burn rate, we bootstrapped and kept the team small: just myself handling Design and Product, while my co-founder, an ex-YC alum with a background in Finance, managed Engineering and Business.

analysing the problem

02 /

I gathered insights by interviewing hedge fund analysts and observing their work firsthand. I spent time in their offices, attended research meetings, and leveraged my co-founder’s domain knowledge. Additionally, I reviewed analyst reports and used tools like Bloomberg, Koyfin, and Alpha Sense to understand their mindset.

analysing the problem

02 /

I gathered insights by interviewing hedge fund analysts and observing their work firsthand. I spent time in their offices, attended research meetings, and leveraged my co-founder’s domain knowledge. Additionally, I reviewed analyst reports and used tools like Bloomberg, Koyfin, and Alpha Sense to understand their mindset.

analysing the problem

02 /

I gathered insights by interviewing hedge fund analysts and observing their work firsthand. I spent time in their offices, attended research meetings, and leveraged my co-founder’s domain knowledge. Additionally, I reviewed analyst reports and used tools like Bloomberg, Koyfin, and Alpha Sense to understand their mindset.

analysing the problem

02 /

I gathered insights by interviewing hedge fund analysts and observing their work firsthand. I spent time in their offices, attended research meetings, and leveraged my co-founder’s domain knowledge. Additionally, I reviewed analyst reports and used tools like Bloomberg, Koyfin, and Alpha Sense to understand their mindset.

AI First web app

03 /

We aimed to build an AI-first web app for stock research, replicating the interactive depth of conversations with a human analyst. Users could "speak" to a virtual CEO, asking questions about the company through a dynamic chat. The app featured a real-time knowledge graph presenting both qualitative and quantitative information from sources like investor sites, earnings calls, social media, and news. AI would synthesise this data to provide its stance on stock prices. The design drew inspiration from Bloomberg Terminal, familiar to most users, but added a futuristic vibe.

AI First web app

03 /

We aimed to build an AI-first web app for stock research, replicating the interactive depth of conversations with a human analyst. Users could "speak" to a virtual CEO, asking questions about the company through a dynamic chat. The app featured a real-time knowledge graph presenting both qualitative and quantitative information from sources like investor sites, earnings calls, social media, and news. AI would synthesise this data to provide its stance on stock prices. The design drew inspiration from Bloomberg Terminal, familiar to most users, but added a futuristic vibe.

AI First web app

03 /

We aimed to build an AI-first web app for stock research, replicating the interactive depth of conversations with a human analyst. Users could "speak" to a virtual CEO, asking questions about the company through a dynamic chat. The app featured a real-time knowledge graph presenting both qualitative and quantitative information from sources like investor sites, earnings calls, social media, and news. AI would synthesise this data to provide its stance on stock prices. The design drew inspiration from Bloomberg Terminal, familiar to most users, but added a futuristic vibe.

AI First web app

03 /

We aimed to build an AI-first web app for stock research, replicating the interactive depth of conversations with a human analyst. Users could "speak" to a virtual CEO, asking questions about the company through a dynamic chat. The app featured a real-time knowledge graph presenting both qualitative and quantitative information from sources like investor sites, earnings calls, social media, and news. AI would synthesise this data to provide its stance on stock prices. The design drew inspiration from Bloomberg Terminal, familiar to most users, but added a futuristic vibe.

Challenges in PMF

04 /

GPT-3.5's frequent hallucinations made it unreliable. While many companies were attempting similar solutions, our limited resources made it impossible to compete with established giants like Bloomberg and Yahoo Finance, who had access to vast data pools. Additionally, the niche hedge fund market we targeted was too small to support significant scaling of the product or business.

Challenges in PMF

04 /

GPT-3.5's frequent hallucinations made it unreliable. While many companies were attempting similar solutions, our limited resources made it impossible to compete with established giants like Bloomberg and Yahoo Finance, who had access to vast data pools. Additionally, the niche hedge fund market we targeted was too small to support significant scaling of the product or business.

Challenges in PMF

04 /

GPT-3.5's frequent hallucinations made it unreliable. While many companies were attempting similar solutions, our limited resources made it impossible to compete with established giants like Bloomberg and Yahoo Finance, who had access to vast data pools. Additionally, the niche hedge fund market we targeted was too small to support significant scaling of the product or business.

Challenges in PMF

04 /

GPT-3.5's frequent hallucinations made it unreliable. While many companies were attempting similar solutions, our limited resources made it impossible to compete with established giants like Bloomberg and Yahoo Finance, who had access to vast data pools. Additionally, the niche hedge fund market we targeted was too small to support significant scaling of the product or business.

Change in Direction

05 /

IR websites often overwhelm investors with dense information, including lengthy earnings call transcripts, complex financial reports, and unstructured data, making it difficult to quickly find key insights. This can cause delays in research and lead to frustration. A chat-based AI could streamline this process by offering direct answers, reducing research time, and providing IR teams with insights into investor queries to enhance communication.

Change in Direction

05 /

IR websites often overwhelm investors with dense information, including lengthy earnings call transcripts, complex financial reports, and unstructured data, making it difficult to quickly find key insights. This can cause delays in research and lead to frustration. A chat-based AI could streamline this process by offering direct answers, reducing research time, and providing IR teams with insights into investor queries to enhance communication.

Change in Direction

05 /

IR websites often overwhelm investors with dense information, including lengthy earnings call transcripts, complex financial reports, and unstructured data, making it difficult to quickly find key insights. This can cause delays in research and lead to frustration. A chat-based AI could streamline this process by offering direct answers, reducing research time, and providing IR teams with insights into investor queries to enhance communication.

Change in Direction

05 /

IR websites often overwhelm investors with dense information, including lengthy earnings call transcripts, complex financial reports, and unstructured data, making it difficult to quickly find key insights. This can cause delays in research and lead to frustration. A chat-based AI could streamline this process by offering direct answers, reducing research time, and providing IR teams with insights into investor queries to enhance communication.

MVP

06 /

We developed Ralina, an AI persona designed for Investor Relations teams, which could be embedded on Investor Relations sites. The system offered answers by linking directly to specific PDFs and pages, facilitating quick access to detailed information. To make Ralina feel more "alive," we incorporated features like blinking and reacting to being tapped. While we had plans to implement emotional responses based on the user's tone, technical challenges limited this feature. For voice interaction, we utilized 11lab's API.

MVP

06 /

We developed Ralina, an AI persona designed for Investor Relations teams, which could be embedded on Investor Relations sites. The system offered answers by linking directly to specific PDFs and pages, facilitating quick access to detailed information. To make Ralina feel more "alive," we incorporated features like blinking and reacting to being tapped. While we had plans to implement emotional responses based on the user's tone, technical challenges limited this feature. For voice interaction, we utilized 11lab's API.

MVP

06 /

We developed Ralina, an AI persona designed for Investor Relations teams, which could be embedded on Investor Relations sites. The system offered answers by linking directly to specific PDFs and pages, facilitating quick access to detailed information. To make Ralina feel more "alive," we incorporated features like blinking and reacting to being tapped. While we had plans to implement emotional responses based on the user's tone, technical challenges limited this feature. For voice interaction, we utilized 11lab's API.

MVP

06 /

We developed Ralina, an AI persona designed for Investor Relations teams, which could be embedded on Investor Relations sites. The system offered answers by linking directly to specific PDFs and pages, facilitating quick access to detailed information. To make Ralina feel more "alive," we incorporated features like blinking and reacting to being tapped. While we had plans to implement emotional responses based on the user's tone, technical challenges limited this feature. For voice interaction, we utilized 11lab's API.

Traction & distribution

07 /

IR teams had minimal budgets and were highly price-sensitive, making it challenging to justify the investment. Additionally, creating an SEC-compliant solution proved to be too complex. Furthermore, venture capitalists were not enthusiastic about a chat-style interface, adding another hurdle to the project's viability.

Traction & distribution

07 /

IR teams had minimal budgets and were highly price-sensitive, making it challenging to justify the investment. Additionally, creating an SEC-compliant solution proved to be too complex. Furthermore, venture capitalists were not enthusiastic about a chat-style interface, adding another hurdle to the project's viability.

Traction & distribution

07 /

IR teams had minimal budgets and were highly price-sensitive, making it challenging to justify the investment. Additionally, creating an SEC-compliant solution proved to be too complex. Furthermore, venture capitalists were not enthusiastic about a chat-style interface, adding another hurdle to the project's viability.

Traction & distribution

07 /

IR teams had minimal budgets and were highly price-sensitive, making it challenging to justify the investment. Additionally, creating an SEC-compliant solution proved to be too complex. Furthermore, venture capitalists were not enthusiastic about a chat-style interface, adding another hurdle to the project's viability.

SENTIMENT-DRIVEN INVESTING

08 /

New investors are often overwhelmed by complex data and financial jargon. Inspired by recent market trends where sentiment-driven communities influenced stock prices—like GameStop—we developed a platform that personifies companies as everyday objects (fruits, flowers, spices) to make investing fun and relatable. I can assure you we weren't losing our minds at this point! Our hypothesis was that this creative approach will engage beginners, helping them feel emotionally connected with the companies they are investing in.

SENTIMENT-DRIVEN INVESTING

08 /

New investors are often overwhelmed by complex data and financial jargon. Inspired by recent market trends where sentiment-driven communities influenced stock prices—like GameStop—we developed a platform that personifies companies as everyday objects (fruits, flowers, spices) to make investing fun and relatable. I can assure you we weren't losing our minds at this point! Our hypothesis was that this creative approach will engage beginners, helping them feel emotionally connected with the companies they are investing in.

SENTIMENT-DRIVEN INVESTING

08 /

New investors are often overwhelmed by complex data and financial jargon. Inspired by recent market trends where sentiment-driven communities influenced stock prices—like GameStop—we developed a platform that personifies companies as everyday objects (fruits, flowers, spices) to make investing fun and relatable. I can assure you we weren't losing our minds at this point! Our hypothesis was that this creative approach will engage beginners, helping them feel emotionally connected with the companies they are investing in.

SENTIMENT-DRIVEN INVESTING

08 /

New investors are often overwhelmed by complex data and financial jargon. Inspired by recent market trends where sentiment-driven communities influenced stock prices—like GameStop—we developed a platform that personifies companies as everyday objects (fruits, flowers, spices) to make investing fun and relatable. I can assure you we weren't losing our minds at this point! Our hypothesis was that this creative approach will engage beginners, helping them feel emotionally connected with the companies they are investing in.

AI AUTOMATION

09 /

We started by experimenting with different prompts to generate dynamic content. Using company names, we crafted AI-driven, humorous one-liners and descriptions inspired by fruits, flowers, and spices. After extensive trial and error, we finally reached a level of content quality we were happy with.

AI AUTOMATION

09 /

We started by experimenting with different prompts to generate dynamic content. Using company names, we crafted AI-driven, humorous one-liners and descriptions inspired by fruits, flowers, and spices. After extensive trial and error, we finally reached a level of content quality we were happy with.

AI AUTOMATION

09 /

We started by experimenting with different prompts to generate dynamic content. Using company names, we crafted AI-driven, humorous one-liners and descriptions inspired by fruits, flowers, and spices. After extensive trial and error, we finally reached a level of content quality we were happy with.

AI AUTOMATION

09 /

We started by experimenting with different prompts to generate dynamic content. Using company names, we crafted AI-driven, humorous one-liners and descriptions inspired by fruits, flowers, and spices. After extensive trial and error, we finally reached a level of content quality we were happy with.

Image Generation

10 /

The initial prompt allowed me to auto-generate images using the DALL-E API, but the results lacked consistency and quality. To tackle this, I manually generated around 300 images in MidJourney to ensure the desired style and quality. Then, I used AI to remove the backgrounds, automating part of the process. Although it was time-consuming, I was very happy with the final results.

Image Generation

10 /

The initial prompt allowed me to auto-generate images using the DALL-E API, but the results lacked consistency and quality. To tackle this, I manually generated around 300 images in MidJourney to ensure the desired style and quality. Then, I used AI to remove the backgrounds, automating part of the process. Although it was time-consuming, I was very happy with the final results.

Image Generation

10 /

The initial prompt allowed me to auto-generate images using the DALL-E API, but the results lacked consistency and quality. To tackle this, I manually generated around 300 images in MidJourney to ensure the desired style and quality. Then, I used AI to remove the backgrounds, automating part of the process. Although it was time-consuming, I was very happy with the final results.

Image Generation

10 /

The initial prompt allowed me to auto-generate images using the DALL-E API, but the results lacked consistency and quality. To tackle this, I manually generated around 300 images in MidJourney to ensure the desired style and quality. Then, I used AI to remove the backgrounds, automating part of the process. Although it was time-consuming, I was very happy with the final results.

Live Site

Implementation

11 /

I built the front end in Framer to easily control the look, feel, and interaction of the site. Using Framer’s CMS, we created dynamic pages and data integration by importing it from a CSV file. Since Framer’s native capabilities weren't sufficient for features like the drop-down menu and search functionality, we custom-coded these elements. To ensure a responsive design, I created numerous breakpoints, maintaining consistent spacing and design across devices.

Implementation

11 /

I built the front end in Framer to easily control the look, feel, and interaction of the site. Using Framer’s CMS, we created dynamic pages and data integration by importing it from a CSV file. Since Framer’s native capabilities weren't sufficient for features like the drop-down menu and search functionality, we custom-coded these elements. To ensure a responsive design, I created numerous breakpoints, maintaining consistent spacing and design across devices.

Implementation

11 /

I built the front end in Framer to easily control the look, feel, and interaction of the site. Using Framer’s CMS, we created dynamic pages and data integration by importing it from a CSV file. Since Framer’s native capabilities weren't sufficient for features like the drop-down menu and search functionality, we custom-coded these elements. To ensure a responsive design, I created numerous breakpoints, maintaining consistent spacing and design across devices.

Implementation

11 /

I built the front end in Framer to easily control the look, feel, and interaction of the site. Using Framer’s CMS, we created dynamic pages and data integration by importing it from a CSV file. Since Framer’s native capabilities weren't sufficient for features like the drop-down menu and search functionality, we custom-coded these elements. To ensure a responsive design, I created numerous breakpoints, maintaining consistent spacing and design across devices.

Learnings

12 /

Running a startup demands wearing many hats, from sales and business to marketing, making flexibility and adaptability essential for managing diverse responsibilities and challenges. In hindsight, we spent too much time initially trying to secure investment instead of focusing on building the product. Prioritising product development from the start would have provided a solid foundation, attracting investment more naturally. With more resources, we would have built the product properly rather than relying on Framer as a platform. This approach would have enabled us to pull in dynamic content, integrate AI more effectively, and expand our feature set with additional features.

Learnings

12 /

Running a startup demands wearing many hats, from sales and business to marketing, making flexibility and adaptability essential for managing diverse responsibilities and challenges. In hindsight, we spent too much time initially trying to secure investment instead of focusing on building the product. Prioritising product development from the start would have provided a solid foundation, attracting investment more naturally. With more resources, we would have built the product properly rather than relying on Framer as a platform. This approach would have enabled us to pull in dynamic content, integrate AI more effectively, and expand our feature set with additional features.

Learnings

12 /

Running a startup demands wearing many hats, from sales and business to marketing, making flexibility and adaptability essential for managing diverse responsibilities and challenges. In hindsight, we spent too much time initially trying to secure investment instead of focusing on building the product. Prioritising product development from the start would have provided a solid foundation, attracting investment more naturally. With more resources, we would have built the product properly rather than relying on Framer as a platform. This approach would have enabled us to pull in dynamic content, integrate AI more effectively, and expand our feature set with additional features.

Learnings

12 /

Running a startup demands wearing many hats, from sales and business to marketing, making flexibility and adaptability essential for managing diverse responsibilities and challenges. In hindsight, we spent too much time initially trying to secure investment instead of focusing on building the product. Prioritising product development from the start would have provided a solid foundation, attracting investment more naturally. With more resources, we would have built the product properly rather than relying on Framer as a platform. This approach would have enabled us to pull in dynamic content, integrate AI more effectively, and expand our feature set with additional features.

Equity research is costly and requires tedious, error-prone data collection.Analysts struggle to cover numerous companies thoroughly, limiting their ability to generate insights and new ideas.

Equity research is costly and requires tedious, error-prone data collection.Analysts struggle to cover numerous companies thoroughly, limiting their ability to generate insights and new ideas.

Our Approach

01 /

Our Approach

01 /

Our Approach

01 /

Our Approach

01 /

To minimize burn rate, we bootstrapped and kept the team small: just myself handling Design and Product, while my co-founder, an ex-YC alum with a background in Finance, managed Engineering and Business.

analysing the problem

02 /

analysing the problem

02 /

analysing the problem

02 /

analysing the problem

02 /

AI First web app

03 /

AI First web app

03 /

AI First web app

03 /

AI First web app

03 /

The concept was that, through prompts, the UI would dynamically generate (GEN UI) dashboards and watchlists, making the interface adaptable to the user needs.

The concept was that, through prompts, the UI would dynamically generate (GEN UI) dashboards and watchlists, making the interface adaptable to the user needs.

Challenges in PMF

04 /

Challenges in PMF

04 /

Challenges in PMF

04 /

Challenges in PMF

04 /

Change in Direction

05 /

Change in Direction

05 /

Change in Direction

05 /

Change in Direction

05 /

MVP

06 /

MVP

06 /

MVP

06 /

MVP

06 /

WE experimented with various concepts, from voice interaction, akin to Siri, to dashboards displaying KPIs for IR teams. I also designed the ability for users tocustom-create their company avatar, allowing them to fine-tune its voice and response style.

I experimented with various concepts, from 3D floating orbs for voice interaction, akin to Siri, to dashboards displaying KPIs for IR teams. I also designed the ability for users to custom-create their company avatar, allowing them to fine-tune its voice and response style.

Traction & distribution

07 /

Traction & distribution

07 /

Traction & distribution

07 /

Traction & distribution

07 /

After spending months trying to sell the product, we were running out of runway, and our burn rate had become unsustainable. As a last attempt, we decided to pivot to KV-01.

After spending months trying to sell the product, we were running out of runway, and our burn rate had become unsustainable. As a last attempt, we decided to pivot to KV-01.

SENTIMENT-DRIVEN INVESTING

08 /

SENTIMENT-DRIVEN INVESTING

08 /

SENTIMENT-DRIVEN INVESTING

08 /

SENTIMENT-DRIVEN INVESTING

08 /

AI AUTOMATION

09 /

AI AUTOMATION

09 /

AI AUTOMATION

09 /

AI AUTOMATION

Equity research is costly and requires tedious, error-prone data collection.

Analysts struggle to cover numerous companies thoroughly, limiting their ability to generate insights and new ideas.

Equity research is costly and requires tedious, error-prone data collection.

Analysts struggle to cover numerous companies thoroughly, limiting their ability to generate insights and new ideas.

WE experimented with various concepts, from voice interaction, akin to Siri, to dashboards displaying KPIs for IR teams.

I also designed the ability for users to
custom-create their company avatar, allowing them to fine-tune its voice and response style.

I experimented with various concepts, from 3D floating orbs for voice interaction, akin to Siri, to dashboards displaying KPIs for IR teams.

I also designed the ability for users to
custom-create their company avatar, allowing them to fine-tune its voice and response style.

Subtle interactions, like slowing the ticker on hover and animating fruits over buttons, were added for a dynamic feel.

I wanted the cards to evoke the excitement of a slot machine each time a new company loaded in.

Subtle interactions, like slowing the ticker on hover and animating fruits over buttons, were added for a dynamic feel.

I wanted the cards to evoke the excitement of a slot machine each time a new company loaded in.

Launched and gained 20-30 MAUs in the first month, focusing on building a social media presence and targeting the meme stock and crypto markets for viral growth.

we quickly realised that sustaining this required a different approach and given more time could have explored it further.

Launched and gained 20-30 MAUs in the first month, focusing on building a social media presence and targeting the meme stock and crypto markets for viral growth.

we quickly realised that sustaining this required a different approach and given more time could have explored it further.