vangax.com

VANG / DEVANG VERMA

ai engineer · undergrad researcher ·

++ main.panel/

music.panel/

++ no track loaded

0:00 0:00

▶

■

RND

VOL

++ ready ||

news.panel/

06.04.2026
Released ERP Explorer — interactive BigP3BCI dataset explorer with live scalp topography.

04.24.2026
Launched Oneiric — generative extension of Amalgam. slime mold that dreams synthetic data.

04.19.2026
Published Amalgam — hybrid digital-biological neural net with slime mold as layer 4.

03.26.2026
ENGRAM v0.2 prototype live — second-brain system for the new tech era.

03.13.2026
Blackwall V1 — 5-phase emergent language detection & containment in multi-agent RL.

info.panel/

mailing list:

++ content.panel

++ home : welcome ||

i'm vang. i'm 19 and i'm trying to build things that probably shouldn't exist yet — and figure out the theory after. somewhere between my first and second year of college i realized that the things i actually wanted to work on weren't in any syllabus, so i started working on them anyway. final year now, 1+ year of research, and i still feel like i'm just getting started.

++ currently.working_on/ ||

Working on a Cybersec Project "Mirage Source" with my childhood friend Vinayak
BCI For Prosthetics — building toward neural interfaces that restore lost function

++ research.panel || interests & timeline

++ research interests ||

building tars: a short timeline

july 2024

the idea becomes real

for a long time, ever since i watched the movie interstellar the idea of tars (an ai robot that helped humans) lived only in my head. in july 2024, i decided to stop thinking about it and actually try building something. there was no plan, no roadmap, and no research goal at this stage. i just wanted to see if i could create a small assistant that worked on my own machine and felt personal.

i started by installing a large language model locally using ollama. the decision to keep everything offline was intentional. i wanted to understand what an ai system feels like when it has no internet, no cloud support, and no shortcuts.

late aug 2024

giving the system a personality

once the system was running locally, i started thinking less about answers and more about behavior. i experimented with ways to control how the system responds, adding adjustable settings for humor, joke frequency, and overall seriousness during conversations.

one feature that felt especially important to me was control. i added a trigger phrase, "hey its serious now tars", which switches the system into a focused mode. in this state, the assistant(tars) ignores casual language and concentrates only on solving the problem i give it.

early oct 2024

learning how humans talk to machines

as i used the system more, typing started to feel limiting. i wanted the interaction to feel closer to how humans naturally communicate. i integrated a speech recognition system so i could talk to him directly.

i also spent time adjusting his voice same as tars from the movie. this made the assistant feel more dependable and less like a toy and it was really fun sometimes because when it was in non focused mode, tars would make funny or unfunny jokes during the work.

march 2025

from experiment to research interest

by march 2025, this project was no longer just something i played with. using the system daily, completely offline, changed how i think about ai i started seeing intelligent agents as tools that could genuinely support humans in environments where connectivity is unreliable or unavailable i started to think about if there's an major outbreak these ai systems can help humans in a way never imagined before.

this experience made me want to take this idea further at a research level. i became interested in how autonomous ai systems can be designed to be controllable, reliable, and helpful, while still remaining flexible enough to adapt to different situations and users.

brain–computer interfaces: curiosity that turned into research

early curiosity

the idea that the brain could speak to machines

one of the ideas that fascinated me early on was the possibility that the human brain could directly communicate with machines. the concept of translating neural signals into computational outputs felt almost like science fiction, yet the scientific field of brain–computer interfaces shows that this idea is already partially real.

i became interested in understanding how electrical activity measured from the brain could be processed, decoded, and interpreted using machine learning algorithms.

2025

building my first neural decoding system

this curiosity eventually led me to work on my first brain–computer interface system using eeg signals and machine learning.

i developed a system that processes neural signals recorded during motor imagery tasks and decodes them using a machine learning pipeline combining signal processing, spatial filtering, and neural networks.

the system extracts features from eeg signals using filter bank common spatial patterns (fbcsp) and then learns a latent representation of neural states using a neural embedding model.

may 2026

eeg-based authentication

my current research moves into a new direction within the eeg space. i am working on a system that uses eeg signals as a basis for user authentication, exploring how neural activity can serve not just as a communication channel but as a unique identifier tied to an individual's cognition.

the work sits at the intersection of cognitive neuroscience, signal processing, and machine learning and is being developed under active research supervision.

the bigger picture

i believe bci will redefine what it means to be human

authentication is a starting point. what i want to build toward is something much larger. i want to work on systems and theories that explore how brain–computer interfaces can be fully utilized to push human capability into territory we have not yet reached.

i think about the people who cannot speak, cannot move, cannot see. i think about what it would mean to give someone back a limb through a neural prosthetic, to restore sight through cortical visual interfaces, to let a paralyzed person navigate the world through thought alone. these are not distant fantasies. they are engineering problems waiting to be solved.

my long-term ambition is to contribute to that work and push human limits of healthy individuals to help build the theoretical and computational foundations that make true neural accessibility possible. not just for some people(healthy ones), for anyone whose brain works but whose body has failed them. that is the version of bci research i want to spend my life on

why astrophysics: a personal timeline

2007–2012

first questions about space

my interest in space started very early. growing up, i was fascinated by movies like zathura the space adventure and games like star trek that showed different worlds, races, and journeys through space. at that age(5), i did not understand the science behind it, but space felt vast and unreal that things like this actually exist!!!

even before i knew what astrophysics meant, i found myself asking simple questions about planets, stars, and what exists beyond earth and i remember annoying my grade 1 class teacher by showing them cool space images and art from a book i had back then and asking her questions about those images all the time

middle school

building space projects

during school exhibitions, my curiosity turned into physical projects. i built models of the solar system and moon's impact on earth's oceans these were simple, but they helped me visualize how planets move and how space is structured.

this was the first time i realized that space is not just something to watch in movies, but something that can be studied and understood.

high school

understanding the science

in high school, my interest in space became more serious through physics. i started learning about gravitational force, motion, rotational motion, orbits, and basic mechanics related to space and planetary movement and modern physics which is my favourite topic in physics.

this phase helped me understand that astrophysics is built on strong physical principles, not imagination alone.

college period

computational astrophysics

when i entered college and started focusing on data science and machine learning, i naturally tried to connect coding with my long-standing interest in space. this is where astronomy and astrophysics began to intersect with programming for me.

i started exploring multiple research paper related to machine learning models and frameworks related to space and space data which naturally led me to start working with and implementing models and frameworks on space data and real time data from satellites which is available publically.

2024–2026

machine learning for astrophysics

over the last few years, i have worked on more than six projects related to astrophysics, astronomy, computational physics, and space science using machine learning and deep learning and neural frameworks - i have tried to understand these concepts on my own

these projects include tasks such as exoplanet,anomaly detection in astronomical data, time-series analysis, and exploratory modeling of space related datasets. through this work, i realized that machine learning can act as a powerful tool to support astrophysical research rather than replace physical understanding.

this journey is why i am deeply interested and would love to pursue astrophysics through a computational and machine learning perspective during my master’s studies and beyond

human–ai interfaces and neuroadaptive systems

inspiration 2022

when science fiction starts asking real questions

many of my ideas about human–machine interaction were inspired by speculative concepts in science fiction ie a game called Cyberpunk 2077, particularly worlds where humans and intelligent systems interact seamlessly. I first played this game in 2022 which amazed me to my core , since then ive always had a wish to live in such a world.

these stories imagine environments where technology can respond directly to human thoughts, emotions, or cognitive states.

while these ideas are fictional, they point toward real research questions:

can intelligent systems adapt dynamically to human neural or cognitive signals? (im sure they can, to a scale we cannot say as of now - 13th march 2026 10:02 am)

exploration 2025

exploring neuroadaptive interaction systems

this question led me to explore the concept of neuroadaptive systems — systems where machine learning models respond dynamically to signals reflecting human mental states.

such systems could combine brain–computer interfaces, machine learning, and interactive systems to create environments where technology adapts to users in real time and help make human mind more efficient.

examples include interfaces that respond to attention levels, cognitive load, or emotional states inferred from neural or physiological signals, these type of interfaces can help humans respond more accurately.

future research

toward adaptive human–machine systems

my long-term research interest in this area focuses on designing systems where humans and intelligent agents can interact in more natural and adaptive ways

this includes studying how machine learning models can integrate signals from human cognition and adapt their behavior accordingly.

i am particularly interested in the intersection of neurotechnology, artificial intelligence, and human–computer interaction, where neural signals could guide intelligent systems and create new forms of human–machine collaboration.

++ reflections.panel || thoughts & transitions

++ reflections ||

not research. not projects. just things i've thought hard about — failures, realizations, the slow shifts in how i see things.

April 2026

thinking to change my focus from core theory

ive been learning physics deeply for the last 5 years and i genuinely love it going into theories, understanding why things happen, thinking about the universe and all that. but recently as im growing i can feel something changing in my head. my mind is not fully satisfied just thinking anymore. its pushing me towards actually making things, applying what ive been learning, building stuff that actually uses that physics or reasons it in some way.

earlier i used to enjoy just sitting and learning, trying to understand concepts again and again. but now i get way more dopamine from building something, even if its small, even if it fails. it just feels more real. like im actually doing something instead of just thinking about doing it.

and at the same time the world is moving very fast ai agents, new tools, people building crazy stuff every day. and when im just stuck in learning theory for long periods it starts feeling like nothing is visible, and then fomo kicks in because i know that in the same time i could have shipped something, experimented more, instead of staying in that loop.

so yeah. im slowly shifting from a thinking phase to a building phase not because i stopped loving physics, but because i want to use it now. i want to turn all that thinking into something real. something that exists outside my head.

March 2026

on engram

engram was never something i planned. it just came into my mind randomly one night while i was thinking and it hit really deep not a normal idea, it felt real and heavy. i was thinking about people who are going to be left behind in this fast moving world. everything is speeding up and at the same time people are getting lonelier day by day.

i was reading the bhagavad gita around that time and one idea stuck with me that once you find your true self you stop feeling suffering the same way. you start observing it from a third perspective, almost like supporting your own material body instead of being controlled by it. that is where engram started forming. what if there was something that reminds you of who you actually are, but in a safe way. not addictive. not replacing you. just bringing you back to yourself again and again.

i started very raw testing llms, writing code manually, no proper structure. took some ui inspiration from cyberpunk because i wanted it to feel a certain way. something minimal but deep. after that i tested the llm with constraints around identity, local memory, soul identity. terms i didnt even fully understand at that time but i kept experimenting.

the ui phase was honestly full of failures. things breaking, layouts not making sense, nothing feeling right. then i shifted to features initially i had 19 in mind but as i started building i realised most were unnecessary. cut it down to 2 core things and focused only on them.

i was working almost 18 hours a day. unhealthy, but it felt like i was getting closer to something real. then the biggest slowdown hallucination problems with the llm, no proper solution, especially keeping everything local. limited context, constant constraints. it felt like i was fighting the system itself.

still i kept going. built an 8 layer recall system, tried to structure memory better, worked on something like a soul system to maintain continuity of identity. experimented with a memory palace wing architecture in early april sounded good in theory, didnt work practically.

right now im still working on it, just slower. the idea never left. it still feels like something important. something that needs to exist. i just need to figure out how to make it real in the right way.

"i was thinking of something way ahead for the current time."

January 2026

upcoming summer

this year started and i came back home for winter holidays and something just felt different. a wave of nostalgia hit me out of nowhere. hometown just has that feeling you cant really explain it just feels right. i went out to play football with my school friends after so long and it felt like nothing had changed, and at the same time everything had changed. i was also spending a lot of time with my family, just sitting, talking, being around them, while also working on nebula in between.

for a moment i genuinely felt like what if i just earn well and settle here. live a simple life around the same people, same places. it felt peaceful in a way i dont feel in my current city. but then again i know i cant just become homesick and stop moving forward. there is more i want to do, more i want to build, more i want to explore.

so i made a decision for myself this summer i want to do things differently. take up a remote job or internship so i can stay at home, wake up in my own space, go out everyday, meet people, play football, spend time with family, and still keep building at the same time. balance, not just work.

also im genuinely excited for the world cup. watching matches with my family at home, that whole vibe just hits different. especially after last world cup when my idol finally won it that moment was unreal. now i really want portugal or brazil to win this one. it just feels like it would complete something in a weird way.

"i think the willingness to enjoy life while doing the karma is the best thing i can do to myself."

2023–2025

what just happened?

i was studying physics and preparing for medical with applied maths during my senior secondary. i was fully into it that was the only path i could see for myself. hours of concepts, problem solving, thinking about how everything connects. mentally invested, not just academically.

then in 2024 during my school farewell there was one moment that stayed with me more than anything else. very emotional everything ending at once. after the whole ceremony when everyone had started leaving, me and my best friend just stayed back on the ground. we didnt really want to leave. so we picked up the ball and started playing football in our formal clothes pants, shirts, everything. we played for more than an hour. talking in between, playing, laughing, but also feeling that heaviness that this is the last time we are here like this. it wasnt even about the game. it was about holding on to that moment a little longer. i never thought i would miss school this much. but that day made me realise it.

but things didnt go the way i expected. i couldnt get into medical and that hit harder than i thought it would not just about not getting in, but because the direction i had in my head just disappeared. during that time i also slowly phased out of physics without even realising it. like everything just paused and i didnt know what to hold on to.

i ended up coming to symbiosis for college. first year was simple learning programming basics from scratch, playing football whenever i could, enjoying small things. but there was a quiet shift happening in how i was thinking. towards the end of 2024 i started getting back into things on my own. not structured, just exploring whatever felt interesting. self-driven and a bit isolated, not really showing it to people, just going with the flow.

then in 2025 i completed my internship at AIT GLOBAL that experience changed a lot for me exposed me to real systems, real workflows, real problems. mr ajeet singh, the delivery head, helped me understand how systems actually work and how to think practically, not just theoretically. after that i entered second year with a completely different mindset. fire to build, to experiment, to push and keep going deeper into research, but now with action, not just thought

i've learned to adapt. finally.

Nov–Dec 2022

the football incident

i was preparing very rigorously for my upcoming nationwide football tournament clusters, where schools from all over the west region of india come to compete. i was training hard but during one session i pushed myself too much and injured my hip. angry at myself because it could have been avoided. still i forced myself to go.

when we reached it was the same venue from 2019 the same school. i remembered how much fun we had there last time. and somehow it actually happened again. day 1 itself i had already made a lot of new friends from different schools. we were staying in school classrooms. honestly how can it get more fun than that being in a school at night. we were doing all kinds of random activities with students from other schools. we even celebrated my best friend nitin's birthday there which made it even more special.

but then the next day we had our match. we gave everything. i played through a lot of pain but we still couldnt win and got knocked out. it hit hard that was the last tournament i ever attended. i was really sad, felt like a complete loser at that moment. but we stayed for the finals and somewhere inside i told myself i wont give up ever.

and in the meantime the 2022 world cup was going on and i was rooting for messi. after i came back home, me and my family watched every argentina game together. when messi finally won it something just shifted in me genuinely changed my mood, gave me a different kind of energy. like things were not as bad as they felt in that moment.

this is me in 2026 writing this i will play again. i will participate in more tournaments than i ever did. ive been busy with everything but i know i will come back to it

this is me in 2026 writing this — i will play again. i will participate in more tournaments than i ever did.

"The only place that 'success' comes before 'work' is in the dictionary."

++ contact.panel || get in touch

++ contact ||

feel free to reach out — research, collaboration, or just to talk.

email/ vang230107@gmail.com

github/ github.com/vangax

linkedin/ linkedin.com/in/vangax

X/ x.com/vangax_champ