Neurotech Builder & Founder

Reimagining how we’re conditioned to function: mentally, physically, consciously.

My story

I grew up with a women’s health condition no doctor could diagnose.

I tried over 11 therapies. What changed my life seven years later was going after the root cause and building my own treatment stack.

That experience is what drives me.

It’s carried me from Hong Kong to a B.S. in Cognitive Science & Entrepreneurship at UCLA, and now to an M.S. in Biomedical Engineering, specializing in neural engineering, at Columbia University.

Along the way, I’ve come to see Eastern and Western medicine not as opposing systems but as two lenses on the same person. Now I’m building the bridge: root-cause care that reads each person’s body and brain, and uses AI to tailor treatment to them.

Work

My mission is to build for the whole person: mentally, physically and consciously.

Mentally

Building closed-loop, multimodal neuromodulation systems that personalize to each brain.

Industry

EEG + TMS

2025

Led an engineering team designing an at-home brain optimization system that pairs consumer EEG with guided TMS, then ran the validation pilot behind it. Also built a RAG chatbot to train clinicians and mined patient records for biomarkers of treatment response.

“Your presentation was exceptional: clear, thoughtful, and impactful. And you’ve made a real contribution to the team. You leave here with our appreciation and respect, and I’m confident you’ll continue to do great things.”
Fred Walke, CEO

“You have demonstrated incredible focus, thoughtfulness, and work ethic during your time here over the summer. Your work will leave a lasting impression that will improve what we do.”
Alexander Ring, Director of Applied Sciences

Product Development & Data Science Intern, Wave Neuroscience ↗

Research

tVNS

2026–current

Built PyTorch models that forecast arousal escalation from HRV, EDA, respiration and pupillometry before a threshold is crossed, feeding a closed-loop vagus nerve stimulation (tVNS) system for real-time stress regulation. SHAP attribution per subject lays the groundwork for personalized neuromodulation.

Research Assistant, Paul Sajda Lab ↗, Columbia

tFUS

2024–25

Built MATLAB data pipelines and ran neuropsychological assessments (ToPF, BVMT, RAVLT) to support transcranial focused ultrasound neuromodulation research in Parkinson’s disease and MCI.

Research Assistant, Kuhn Lab, UCLA

Community

NeuroGum

Hosting pop-ups, running giveaways and creating content.

Brand Ambassador, NeuroGum ↗

Chandni interviewing a founder beside a SQAI wearable training suit

NEUROHACK NYC

A first-of-its-kind community and event series that helps everyday people track, optimize and neurohack their brains.

Founder · Newly launched

Chandni co-moderating the Future of Focus panel at UCLA Mattel Children’s Hospital

UCLA’s first neurotech hackathon

2024–25

Launched NEUROHACK, UCLA’s first neurotech hackathon centered around “Future of Focus”, sponsored by EMOTIV and NeuroGum.

Lead Project Manager, CruX Neurotechnology at UCLA

Physically

Building intelligent wearables, medical devices & accessible fitness tools that read the body and act on it.

Refuel

2025–26

An app and platform that turns empty studio spots into fitness opportunities by dynamically pricing unbooked class spaces. Launched with Beyond Fitness LA and Lagree by Jenne.

Co-founder, Refuel Movement

Reflex

2025

An AI-powered platform and EMG-integrated apparel that helps users detect muscular imbalances in real time, prevent injury and share movement insights with clinicians or coaches.

1st place, AI x Healthcare Product Space · See Reflex ↗

Projects

Line chart of blood glucose after a meal: a diabetic model with no insulin rises from 163 to 205 mg/dL, while the diabetic model with the closed-loop controller stays near 100 mg/dL, inside the healthy range alongside a healthy-body model

Closed-loop insulin

I built a glucometer from a test strip, an op-amp circuit and an Arduino, then a closed-loop PID controller (P 0.04, I 0.0009, D 0.1) that doses insulin on its own, holding a diabetic model in the healthy range through a meal.

Peak glucose 205 → 113 mg/dL

Scatter plot of ECG model accuracy against training time for six settings, with the chosen 60/40 split and 64 filters highlighted at 83.1 percent in 3.3 minutes

ECG classification

A CNN trained on 27,765 raw ECGs to spot heart attacks, conduction problems, ST changes and hypertrophy, then tuned for the most accuracy per minute of training.

83% accuracy in 3.3 min of training

Bar chart of worst-case thermometer error: 14.9 degrees raw, 4.5 with a 100-point average, 2.9 with a 10-point average, and 1.45 with a 10-point average plus precision resistors

Precision thermometer

I built a thermistor thermometer on a Wheatstone bridge with Steinhart-Hart calibration, then hunted down its error in a real water bath. Averaging tamed the noise, but only precision resistors fixed the offset.

Worst-case error 14.9 → 1.45 °C

Consciously

I advocate for a world where we live with intention: in our consumption, mindsets, careers and role on Earth.

Lives Without Knives

2019–current

Founded a social enterprise connecting plant-based, local and sustainable brands to untapped markets. Built through large-scale festivals, markets and vegan starter kits, with a team of 30+ behind it.

Founder & Executive Director ↗

Chandni speaking on a panel, microphone in hand

LA Climate Week

Panelist

Hashtag Impact podcast cover: Fueling Longevity and Transforming Food Systems, with Abigail Coleman and Chandni Sacheti

Guest Host

Hashtag Impact Podcast: fueling longevity and transforming food systems. I interviewed Abigail Coleman, COO and Co-Founder of Blue Zones Kitchen, on how the Blue Zones philosophy shapes its mission of empowering people to make conscious, healthy choices with ease.

ReThink HK

2024

Built AI-powered tools, a personalized insights engine and a sales calculator, for ReThink HK, Asia’s business conference-expo for climate action and sustainable development, to strengthen client strategy and conference technology.

Sales & AI Strategy Intern ↗

Effective Altruism

2023

Led branding and marketing for Effective Altruism at UCLA, launched and hosted an EA alumni showcase podcast on mission-driven careers, and drove outreach through partnerships with other clubs and EA chapters.

Marketing & Podcast Intern, Effective Altruism at UCLA

Beyond

Asian Hall of Fame

2026

Asian Hall of Fame Heritage Gala, Biltmore Los Angeles.

Model

Tour guide

2025–26

Showing prospective Bruins around UCLA.

Campus Tour Guide, UCLA

Off the clock

A long grazing board with cheeses, fruit, dips and crackers, with watermelon skewers and cupcakes behind it

Cooking & Hosting

From grazing boards to six-course Thai dinners for the people I love.

@theplatingpalette ↗

Three hikers wading upstream through a turquoise river below tall red canyon walls

Hiking & Waterfalls

Waterfall hikes, camping and a mission to visit every national park.

Chandni stamping lettering into the base of a clay bowl at a pottery studio

Always a Beginner

I always keep something going where I’m the worst in the room, whether it’s pottery, boxing, self-defense or salsa dancing.

Skills

Biosignals

EEG (IAF, PSD) · EMG · ECG · HRV · EDA · respiration · pupillometry · bandpass filtering · FFT

Neuromodulation

TMS · closed-loop tVNS · tFUS research · neuropsychological testing (ToPF, BVMT, RAVLT)

Machine & deep learning

PyTorch · 1-D CNNs · temporal convolutional networks · EEGNet · LSTMs · transformers · random forests · SVMs · time-series forecasting · leave-one-subject-out validation · per-subject feature attribution (SHAP) · reinforcement learning · RAG with GPT-4 and Pinecone

Code & tools

Python · MATLAB · R · C++ · SQL · NumPy · pandas · SciPy · scikit-learn · Matplotlib · MNE-Python · Braindecode · MOABB · Weights & Biases · Jupyter · Google Colab (GPU) · Git · Arduino