1. Home
  2. Projects

My projects · Nextentra & collaborations

The collection.

My own projects and collaborations with Nextentra — catalogued like specimens in a field journal.

Project collection

About the collaborations: Nextentra is a technology startup founded by my brother, Arindam Gahlot, a software engineer. These projects were built by Arindam and the Nextentra team — each card credits its builders and notes my role.

Specimen No. 01
My projectLive

Music Discord Bot

A music-focused Discord bot that now serves more than 1,400 servers and has grown into one of the top 25 music bots worldwide.

My project · My role: Product, community & growth

  • Discord
  • Community growth
  • User feedback
Specimen No. 01 · My project

Music Discord Bot

A music-focused Discord bot that now serves more than 1,400 servers and has grown into one of the top 25 music bots worldwide.

My project · My role: Product, community & growth

The soil

Discord communities want an easy, reliable way to listen to music together.

What was planted

I ran the bot as a product: I set feature priorities from what server owners and members asked for, supported the community and grew its reach one server at a time.

What grew

Now serves 1,400+ servers and ranks among the top 25 music bots worldwide.

What I learned

Hands-on experience with technology, digital communities, growth, and understanding what users want from an online service.

Specimen No. 02
NextentraBuilt

Crypto Sniper

A real-time token detection system — a Python pipeline that detects, scores and alerts on newly launched crypto tokens.

Built by Nextentra · My role: Market research & alert testing

  • Python
  • Data pipeline
  • Scoring & alerts
Specimen No. 02 · Nextentra

Crypto Sniper

A real-time token detection system — a Python pipeline that detects, scores and alerts on newly launched crypto tokens.

Built by Nextentra · My role: Market research & alert testing

The soil

New crypto tokens launch constantly, and good signals are buried in noise — false alerts are costly.

What was planted

A modular Python ingestion, scoring and alerting pipeline with automated validation, retries, structured logging and dashboard-ready outputs. I researched how new token launches behave and checked alerts against real launches to flag false positives.

What grew

98% detection accuracy, 96% precision and 94% recall on holdout validation, with fewer false alerts through iterative scoring.

What I learned

How data-driven signals and risk management apply to fast-moving, high-risk markets — and why a false alarm can cost as much as a missed one.

Specimen No. 03
NextentraBuilt

AI Stock Predictor

A machine-learning model that forecasts stock price trends using Python and pandas.

Built by Nextentra · My role: Market research & results review

  • Python
  • pandas
  • Machine learning
Specimen No. 03 · Nextentra

AI Stock Predictor

A machine-learning model that forecasts stock price trends using Python and pandas.

Built by Nextentra · My role: Market research & results review

The soil

Investors want an evidence-based view of where prices might be heading.

What was planted

A machine-learning model trained on historical price data to forecast trends, built with Python and pandas. I helped choose which stocks to follow and compared the model's forecasts with how prices actually moved.

What grew

A working forecasting model that turns years of price history into trend signals for research.

What I learned

Forecasts are probabilities, not promises. Data can sharpen an investment decision, but it can't replace judgment and risk management.

Specimen No. 04
NextentraLive

CharacterPersona

A personality platform with Big Five, attachment-style and love-language assessments, 12 persona archetypes and shareable results.

Built by Nextentra · My role: Product testing & user feedback

  • Web app
  • Big Five research
  • Consumer product
Visit the site
Specimen No. 04 · Nextentra

CharacterPersona

A personality platform with Big Five, attachment-style and love-language assessments, 12 persona archetypes and shareable results.

Built by Nextentra · My role: Product testing & user feedback

The soil

Personality tests are often long, locked behind sign-ups and forgettable — people want results that feel personal and are easy to share.

What was planted

Four free assessments: the flagship 30-question Big Five (about five minutes), a 10-question quick read, attachment style and love languages. Results map to one of 12 archetypes, such as the Architect, the Catalyst and the Strategist. There's no sign-up, and answers stay on the visitor's own device. I tested each assessment as a first-time user and gave feedback on the questions, the results and how easy they are to share.

What grew

Live at characterpersona.com, with four assessments, 12 archetypes and results anyone can share with a single link.

What I learned

What makes a consumer product worth sharing: it's quick, the result feels personal, there's no friction like a sign-up, and there's a reason to send it to a friend.

Specimen No. 05
NextentraLive

FictionPersona

A scenario-led identity reading: a few minutes of instinct-and-pressure scenarios that reveal an original, collectible persona card, ranked by rarity.

Built by Nextentra · My role: Concept & market research

  • Web app
  • Scenario-based quiz
  • Collectible cards
Visit the site
Specimen No. 05 · Nextentra

FictionPersona

A scenario-led identity reading: a few minutes of instinct-and-pressure scenarios that reveal an original, collectible persona card, ranked by rarity.

Built by Nextentra · My role: Concept & market research

The soil

Most personality quizzes ask people to rate themselves, so they answer as who they want to be rather than who they are.

What was planted

A 3–7 minute reading built on scenarios instead of self-ratings. It traces how you lead, protect and hold your ground under pressure, then gives you an original persona card with a rarity from Common to Mythic. New chapters drop weekly; the current one is "Origin Echoes". I helped shape the concept and researched what makes collectible, shareable results appealing.

What grew

Live at fictionpersona.com with a weekly chapter format. No account is needed, and every reading gives exactly one card to keep or share.

What I learned

How rarity, collecting and a weekly release rhythm keep people coming back — the same ideas behind trading cards and limited drops.

Journal

Notes from the semester.

Each post has 300+ words of writing, links to relevant organizations and citations for any media used.

  1. Sept 2026

    My Website & Header Image

    The ideas behind this portfolio and its 3D grass field.