How CoffeeSpace Powers Its Tinder-Like Cofounder Matching App with Proxycurl

We at Proxycurl have had the privilege of speaking with Hazim, the CEO and co-founder of CoffeeSpace, a Tinder/Hinge-like platform for co-founder matching.

Intrigued? We’re here to tell you all about it: its origins, core idea and features, challenges faced, and the tech stack that powers its data-matching algorithm.

The space that CoffeeSpace is delving into is a relatively new and innovatively-disruptive space, insofar that they earned a spot in TechCrunch Disrupt 2024’s Startup Battlefield 200, a premier startup competition by TechCrunch.

Here’s CoffeeSpace’s elevator pitch: Similar to dating apps, users/co-founders can swipe left or right through profiles on the platform. A match occurs when two co-founders swipe right on each other. They can then start chatting and connecting.

Still intrigued? Keep reading.

This was how CoffeeSpace started

Hazim was a finance graduate from MIT with experience in prominent financial roles like the World Bank, yet he always had a passion for startups and innovation.

During the peak of the Covid pandemic, Hazim, like all of us, found himself with a sudden availability of ample time at home. It was during the same period that he met Carin, now the co-founder and CTO of CoffeeSpace. They had countless hours and days of brainstorming and realized there was a significant gap in the co-founder matching space, stemming from their own challenges in finding compatible co-founder partners.

And so, CoffeeSpace was born.

CoffeeSpace’s core idea and features

CoffeeSpace stands out as a disruptive and innovative solution to this notorious and “age-old” struggle within the startup scene of finding co-founders. Traditionally, co-founder matching often involves organizing meetups, or tapping on a community’s existing network such as Y Combinator’s own co‑founder matching network. These meetings are however not easily discoverable, necessitate in-person meet-ups limited to certain cities, and are not scalable.

This is exactly the gap that CoffeeSpace strives to solve.

CoffeeSpace drew major inspiration from popular dating apps like Tinder and Hinge and deployed the familiar "swipe right" feature and algorithms to match users to potential co-founders. They aim to streamline the entire process of finding and connecting with potential co-founders onto a single app platform.

Users can download the app for free. Upon creating a new account, users are prompted to complete a series of quick questionnaires that help the algorithm construct the users’ co-founder profile. This profile will then be displayed and used to match with other users/co-founders.

Here are some snapshots of the questionnaire interface:

How does Proxycurl power CoffeeSpace behind the scenes?

Proxycurl is a full-fledged data enrichment API provider, where developers can pull data at scale about people, companies, contact info, jobs, and more. Our APIs and dataset products power many such applications like CoffeeSpace.

In the series of screenshots shared above, the very first question that CoffeeSpace asks of users during onboarding is their LinkedIn profile URL. CoffeeSpace then uses Proxycurl’s Person Profile Endpoint to enrich the users’ profiles with their educational backgrounds and work experiences automatically. These all happen behind the scenes automatically and at scale for all users during onboarding.

Using Bill Gates as an example. Say he’s looking for a co-founder, he proceeded to download CoffeeSpace and register for an account, this is likely CoffeeSpace’s code input that pulls Bill Gates’ enriched data programmatically via Proxycurl’s People API.

import requests

api_key = 'YOUR_API_KEY'
headers = {'Authorization': 'Bearer ' + api_key}
api_endpoint = 'https://nubela.co/proxycurl/api/v2/linkedin'
params = {
    'linkedin_profile_url': 'https://www.linkedin.com/in/williamhgates/',
    'extra': 'include',
    'github_profile_id': 'include',
    'facebook_profile_id': 'include',
    'twitter_profile_id': 'include',
    'personal_contact_number': 'include',
    'personal_email': 'include',
    'inferred_salary': 'include',
    'skills': 'include',
    'use_cache': 'if-present',
    'fallback_to_cache': 'on-error',
}
response = requests.get(api_endpoint,
                        params=params,
                        headers=headers)

The above code includes other parameters that aren’t directly related to CoffeeSpace’s use case, but it shows the various parameters that can be used with Proxycurl’s Person Profile Endpoint.

And this is the response - specific to CoffeeSpace’s use case - the users’ LinkedIn profile picture, education background and work experience:

{
    "public_identifier": "williamhgates",
    "profile_pic_url": "https://s3.us-west-000.backblazeb2.com/proxycurl/person/williamhgates/profile?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=0004d7f56a0400b0000000001%2F20240927%2Fus-west-000%2Fs3%2Faws4_request&X-Amz-Date=20240927T050405Z&X-Amz-Expires=3600&X-Amz-SignedHeaders=host&X-Amz-Signature=f3959b437dc80e9281bb864bbad28f53312565ab1cd7776bfd6e8e17cea4dc90",
    "experiences": [
        {
            "starts_at": {
                "day": 1,
                "month": 1,
                "year": 2000
            },
            "ends_at": null,
            "company": "Bill & Melinda Gates Foundation",
            "company_linkedin_profile_url": "https://www.linkedin.com/company/bill-&-melinda-gates-foundation",
            "company_facebook_profile_url": null,
            "title": "Co-chair",
            "description": null,
            "location": null,
            "logo_url": "https://s3.us-west-000.backblazeb2.com/proxycurl/company/bill-%26-melinda-gates-foundation/profile?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=0004d7f56a0400b0000000001%2F20240927%2Fus-west-000%2Fs3%2Faws4_request&X-Amz-Date=20240927T050405Z&X-Amz-Expires=1800&X-Amz-SignedHeaders=host&X-Amz-Signature=77937a411477b30790ce699266efabd2cca85edea619f2af747cb5e0f9697884"
        },
        {
            "starts_at": {
                "day": 1,
                "month": 1,
                "year": 2015
            },
            "ends_at": null,
            "company": "Breakthrough Energy ",
            "company_linkedin_profile_url": "https://www.linkedin.com/company/breakthrough-energy",
            "company_facebook_profile_url": null,
            "title": "Founder",
            "description": null,
            "location": null,
            "logo_url": "https://s3.us-west-000.backblazeb2.com/proxycurl/company/breakthrough-energy/profile?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=0004d7f56a0400b0000000001%2F20240927%2Fus-west-000%2Fs3%2Faws4_request&X-Amz-Date=20240927T050405Z&X-Amz-Expires=1800&X-Amz-SignedHeaders=host&X-Amz-Signature=92e52762e76f15b03c35704c35a74b9667c06be40019789485423b169d7a9724"
        },
        {
            "starts_at": {
                "day": 1,
                "month": 1,
                "year": 1975
            },
            "ends_at": null,
            "company": "Microsoft",
            "company_linkedin_profile_url": "https://www.linkedin.com/company/microsoft",
            "company_facebook_profile_url": null,
            "title": "Co-founder",
            "description": null,
            "location": null,
            "logo_url": "https://s3.us-west-000.backblazeb2.com/proxycurl/company/microsoft/profile?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=0004d7f56a0400b0000000001%2F20240927%2Fus-west-000%2Fs3%2Faws4_request&X-Amz-Date=20240927T050405Z&X-Amz-Expires=1800&X-Amz-SignedHeaders=host&X-Amz-Signature=6436aebfa6d4e4aaf00ca41d011ff9011e0985006c5697015617cf4d1ded37ad"
        }
    ],
    "education": [
        {
            "starts_at": {
                "day": 1,
                "month": 1,
                "year": 1973
            },
            "ends_at": {
                "day": 31,
                "month": 12,
                "year": 1975
            },
            "field_of_study": null,
            "degree_name": null,
            "school": "Harvard University",
            "school_linkedin_profile_url": "https://www.linkedin.com/company/1646/",
            "school_facebook_profile_url": null,
            "description": null,
            "logo_url": "https://s3.us-west-000.backblazeb2.com/proxycurl/company/harvard-university/profile?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=0004d7f56a0400b0000000001%2F20240927%2Fus-west-000%2Fs3%2Faws4_request&X-Amz-Date=20240927T050405Z&X-Amz-Expires=1800&X-Amz-SignedHeaders=host&X-Amz-Signature=f364c50c234e9eb9f06a39c3c92a6a5331ca30096f57f405da5472ce3bf39894",
            "grade": null,
            "activities_and_societies": null
        },
        {
            "starts_at": null,
            "ends_at": null,
            "field_of_study": null,
            "degree_name": null,
            "school": "Lakeside School",
            "school_linkedin_profile_url": "https://www.linkedin.com/company/30288/",
            "school_facebook_profile_url": null,
            "description": null,
            "logo_url": "https://s3.us-west-000.backblazeb2.com/proxycurl/company/lakeside-school/profile?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=0004d7f56a0400b0000000001%2F20240927%2Fus-west-000%2Fs3%2Faws4_request&X-Amz-Date=20240927T050405Z&X-Amz-Expires=1800&X-Amz-SignedHeaders=host&X-Amz-Signature=4ef24eca69fdada887d5faa63ceb5606718be833439c930e75d8813638f7f157",
            "grade": null,
            "activities_and_societies": null
        }
    ],
   }

The data retrieved using the Person Profile Endpoint also plays a key role in providing CoffeeSpace with insights into their most prevalent co-founder demographics, informing them about their ideal customer profiles (ICPs) and more, quoting Hazim,

Stanford is our number one user pool by school.

With this information, they know better who their key ICPs are.

In addition to front-end features, Hazim leverages Proxycurl’s Person Profile Endpoint to access data about new users for background verification purposes. This proves especially useful when users provide incomplete information or when their LinkedIn profiles are missing crucial details. In such cases, he retrieves users' data from alternative sources using Proxycurl's API endpoint.

What if users uninstall the app after a successful match?

During the interview, a pressing question was on our minds: Will most users uninstall the app after a successful match? We were glad Hazim gave us a great response to this question.

The short answer is, no. This is because CoffeeSpace’s use case extends beyond just co-founder matching. It also caters to early-stage hiring and networking within the startup scenes as founders continue to find the need to seek talents for their companies.

Even if a successful match is made, co-founders may find that they are not compatible in real life and may need to search for another like-minded partner, again using CoffeeSpace.

Some statistics:

• The average lifespan of a startup is typically 6-12 months, indicating that users may return to CoffeeSpace to find new co-founders for other ideas.
• Early-stage hiring is crucial to the survival of startups (citation), indicating matched co-founders might still use CoffeeSpace to find talents

In short, users do stick around even after successful co-founder matching.

That’s exciting. What’s next for CoffeeSpace’s future?

Improving the cofounder matching algorithm

CoffeeSpace’s superpower lies in its matching algorithm, and that remained the top priority in their product roadmap, to continually improve the algorithm.

One strategy involves leveraging LLM solutions like OpenAI to enhance semantic understanding. This includes interpreting user input, preferences, and enriched profile data to better grasp user personas and enhance matching accuracy.

Besides drawing inspiration from popular dating apps like Tinder and Hinge, CoffeeSpace is also getting inspiration from major social networking sites like TikTok on how to improve its algorithm. Improvements involving tracking various users’ behaviors such as profile views, time spent on profiles, and scrolling habits.

Beyond person data, CoffeeSpace could also pull company data related to their users to enhance that part of their algorithm relating to the users’ professional backgrounds. For instance, with Proxycurl’s Company API, a plethora of additional data points such as a company’s funding history, acquisition information, company size, and more can be pulled to improve user matching.

Expanding into the VC and startup investment matching space

CoffeeSpace has already been attracting interest from investment firms and venture capitalists due to the valuable data insights it possesses on its extensive user base of really-early stage co-founders.

It is common for investment companies to rely on investment signals such as tracking the movements of potential startup founders to identify investment prospects. However, investors will know of this information only after a founder founded a startup and updated his/her social profiles. In some cases like stealth startups, they won’t even know.

Through CoffeeSpace however, VCs can tap into the huge reservoir of potential startup founders even before they start a company, even as early as when they’re exploring ideas or finding a co-founder on CoffeeSpace’s platform.

To better algorithm, data, and co-founder matching

We are stoked to be part of CoffeeSpace’s remarkable growth. Launched in March 2024, CoffeeSpace now has a whooping 5600 users (and growing) and secured $500k in funding.

Similar to the evolution seen in the dating app industry, there initially was a lot of stigma around such tech and algorithms in matching people for love, and now for starting a company. The skepticism was that it’s just odd for tech to connect people for such an intimate and complex endeavor as love, or in this case the startup journey. Over time though, dating apps have become the norm. We believe it’ll happen for the work that CoffeeSpace is doing here too.