Back

Humyn,
Research Assistant for UX Teams.

Humyn,
Research Assistant for UX Teams.

10 Mon.

10 Mon.

from concept to live

10K+

10K+

Analyzed conversations

0→1

0→1

end-to-end ownership

Role

Founder

Role

Founder

Role

Founder

Project Type

Startup

Project Type

Startup

Project Type

Startup

Timeline

02.25 - Present

Timeline

02.25 - Present

Timeline

02.25 - Present

View live site

View live site

View live site

STACK

The brief

How can teams understand real human sentiment at scale without losing nuance?

How can teams understand real human sentiment at scale without losing nuance?

UX and product teams rely on qualitative insight, but analyzing thousands of real conversations is slow, manual, and inconsistent.


Humyn explores how AI can make qualitative research fast, trustworthy, and usable.

  • “Netnography… uses information in online forums to identify and understand consumer needs and decision influences.

    Robert V. Kozinets

  • “Every day, billions talk on social media… This information is a gold mine for consumer-facing industries.”

    MIT Sloan Management Review

  • “User-generated discussions are a rich source for understanding customer behavior and needs.”

    Journal of Marketing Research

  • “Analyze user-forum posts. User forums are sources for important questions and answers that solve problems.

    Nielsen Norman Group

  • “Reddit posts are valuable for search engines and AI… ‘colloquial words about pretty much every topic.’”

    RanjithSteve Huffman, Reddit CEO Kumaran

Context and insight

I spent hours digging through reddit threads, youtube comments, discord chats, and X posts.

Insight existed. Scale did not.

I spent hours digging through reddit threads, youtube comments, discord chats, and X posts.

Insight existed. Scale did not.

Reddit is ranked #1 in trust for product research

Reddit is ranked #1 in trust for product research

The bottleneck is not creativity.
It is understanding people at scale.

The bottleneck is not creativity.
It is understanding people at scale.

Teams track metrics in real time, but qualitative insight lags by days or weeks.

When interpretation is rushed, nuance is lost and confidence drops.

The Problem

Today’s qualitative workflows do not scale.

Today’s qualitative workflows do not scale.

Researchers jump between tabs, notes, spreadsheets, and screenshots to answer basic questions:

  • What are users actually frustrated by?

  • How strong is this sentiment?

  • Is this an edge case or a pattern?


This slows research velocity and weakens product decisions.

Researchers jump between tabs, notes, spreadsheets, and screenshots to answer basic questions:

  • What are users actually frustrated by?

  • How strong is this sentiment?

  • Is this an edge case or a pattern?

This slows research velocity and weakens product decisions.

Solution and product

Humyn turns large volumes of conversation into clear insight.

Humyn turns large volumes of conversation into clear insight.

Humyn analyzes 10,000 plus real conversations and surfaces:

  • Themes

  • Sentiment trends

  • Representative quotes


All while preserving original context and human nuance.


Humyn started as a solo project. With no initial buy in, I built and shipped a working MVP and iterated for six months. The prototype validated real use cases, proved feasibility, and attracted engineers and designers to continue development.

Humyn analyzes 10,000 plus real conversations and surfaces:

  • Themes

  • Sentiment trends

  • Representative quotes


All while preserving original context and human nuance.


Humyn started as a solo project. With no initial buy in, I built and shipped a working MVP and iterated for six months. The prototype validated real use cases, proved feasibility, and attracted engineers and designers to continue development.

Humyn analyzes 10,000 plus real conversations and surfaces:

  • Themes

  • Sentiment trends

  • Representative quotes


All while preserving original context and human nuance.


Humyn started as a solo project. With no initial buy in, I built and shipped a working MVP and iterated for six months. The prototype validated real use cases, proved feasibility, and attracted engineers and designers to continue development.

Icon

View Thousands of Comments

Actionable Insights

Visual Sandbox for Designers

Dashboard
Dashboard
Dashboard
Icon

View Thousands of Comments

Actionable Insights

Visual Sandbox for Designers

Dashboard
Dashboard
Dashboard
Icon

View Thousands of Comments

Actionable Insights

Visual Sandbox for Designers

Dashboard
Dashboard
Dashboard

View live site

View live site

View live site

Everything you need to turn conversations into clarity.

Everything you need to turn conversations into clarity.

Project Snapshot

Quick status: comments ingested, latest sync, highlight summaries, runs processed.

Add Comments

Batch threads with topic + keywords to guide relevance.

Explore with links

Skimmable summaries, topics, and aspects, each with a link back to the original Reddit comment.

Metrics

Overall sentiment and stacked sentiment by aspect to see what drives reactions.

AI Chat

Natural questions, cited answers, one-click jump to Explore.

Design and constraints

From Team to System.

From Team to System.

Clarity over decoration.

  • Low fidelity flows exposed friction early

  • High fidelity designs were shaped by real engineering limits

  • Minimal UI built with shadcn/ui

  • Every screen tied back to a single research question

If it does not answer the core question, it is bloat

If it is slow, static, or inaccurate, trust breaks

Brand, growth and launch

Humyn’s brand was designed to feel intelligent and empathetic.

Humyn’s brand was designed to feel intelligent and empathetic.

Hue the penguin represents curiosity, depth, and human understanding.

Hue the penguin represents curiosity, depth, and human understanding.

I led growth through product driven storytelling and community feedback loops:

  • 2,000 plus followers across social platforms

  • 500 plus Discord testers

  • 500 plus waitlist signups


Goal
Ship clean. Reduce risk. Enable iteration.

Follow our socials here!

Discord

Discord

Discord

Tiktok

Tiktok

Youtube

Youtube

Youtube

X/Twitter

X/Twitter

X/Twitter

Instagram

Instagram

Instagram

Linkedin

Linkedin

Linkedin

Threads

Threads

Threads

Docs

Docs

Docs

Current state & reflection

Humyn is live in closed beta.

Current focus:

  • Real time pipelines and API performance

  • Onboarding and report sharing

  • Guided ML pruning

  • Agentic research workflow


Humyn started as an experiment and became a system.


I learned how to identify the real problem, design under uncertainty, and ship end to end.


Progress comes from turning ambiguity into structure.

More to come.

Current focus:

  • Real time pipelines and API performance

  • Onboarding and report sharing

  • Guided ML pruning

  • Agentic research workflows


Humyn started as an experiment and became a system.


I learned how to identify the real problem, design under uncertainty, and ship end to end.


Progress comes from turning ambiguity into structure.

More to come.

Give us Feedback!

Give us Feedback!

Give us Feedback!

Check out our info Docs.

Check out our info Docs.

Check out our info Docs.

View More

Labor Intelligence

B2B SaaS

Labor Intelligence

B2B SaaS

Labor Intelligence

B2B SaaS

Learn in Context

Figma Plugin

Learn in Context

Figma Plugin

Learn in Context

Figbud (0->1)