Portfolio — 2026 · Cambridge, MA

Finn Pounds

Welcome. A selection of my work in applied AI and data — the projects I've built, the people I've built them with, and a little of what I do outside of it.

M.S. in Analytics, 2026 — Responsible AI & Applied Machine Intelligence
B.S. in Mathematics & Physics — Tufts University, 2019
MBAA-certified packaging operator — brewery fix-it guy, SOP writer, QC-panel designer
Former bike-shop technician
Professional kendama player, co-founder of Paradox, & international freestyle judge
Deeply enjoys a project

View work GitHub

Python · TypeScript · Claude API · PyTorch · YOLOv8 · FastAPI · Next.js

fig. 1 a diffeomorphism of the unit square

fig. 2 — configurations of two and three circles

after Kandinsky · Several Circles, 1926

01 / Selected work

Applied AI, computer vision, and a kendama camera operator.

Things I've built end-to-end — production LLM apps, CV pipelines, and Responsible-AI tooling. A few aren't public yet; links land where they can.

P-01

KendamaVision

End-to-end computer vision for kendama: a fine-tuned YOLOv8 detector (val mAP50 0.91) + BoT-SORT tracker, wrapped in a physics-aware post-processor — gravity priors, depth-normalized gates, a soft string constraint — that recovers the fast, blurry tama the detector misses, lifting per-frame coverage from 67% to 94%, then auto-keyframes a "camera-follows-your-trick" clip. A SAM2 auto-labeling → retrain flywheel keeps sharpening it. The writeup has interactive demos running on real tracking data.

YOLOv8BoT-SORTSAM2Streamlit

Demos & writeup ↗ Live app ↗

P-02

ConceptShape — Semantic Geometry Explorer

An interactive 3D "microscope" for the inside of a language model: watch a sentence's tokens move and cluster layer by layer through the residual stream, compare different models' geometry (CKA + Procrustes), and measure the topology of a corpus with persistent homology — a hands-on take on the Platonic Representation Hypothesis. TransformerLens + PyTorch backend, Next.js + React-Three-Fiber front-end.

TransformerLensFastAPINext.jsR3F

Live demo ↗

P-03

SkyWatch

A production Claude-API weather monitor. An hourly Python pipeline ingests Open-Meteo, NWS, and HRRR GRIB2 data into derived meteorological variables; Claude turns them into plain-language briefings, with response caching and cost-stratified model selection across Haiku / Sonnet / Opus.

Claude APIFastAPINext.jsGRIB2

Live demo ↗

P-04

biasface

A demographic-bias audit of off-the-shelf face models: a linear probe recovers race from ArcFace and ViT-Base embeddings at ~0.85–0.90 held-out AUC — far above chance, though neither was trained to encode it. Paired with the compliance stakes under GDPR, CCPA, and the NIST AI RMF. Built for EAI 6400 (AI Ethics & Governance).

PyTorchResponsible AINIST AI RMF

Read the writeup ↗

P-05

Encrypted Vitals

Can a model flag a tumor from a patient's vitals without ever seeing them in the clear? A logistic-regression classifier is compiled to a fully-homomorphic-encryption circuit (Concrete ML / Zama) and runs on encrypted vitals — the server computes on ciphertext it can't read, and the encrypted predictions match the plaintext ones bit-for-bit. Built for EAI 6400.

Concrete MLFHEscikit-learn

Read the writeup ↗

P-06

GroundTruth

A browser game: classify EuroSAT satellite tiles and race a ResNet18 running live in your browser via ONNX Runtime (WebAssembly) — human vs. AI accuracy, entirely client-side, no server. Built for EAI 6010.

ONNX RuntimeResNet18ReactEuroSAT

Play ↗

More on GitHub

Public repositories, experiments, and works in progress.

@finnpounds

Browse ↗

02 / Academic

Mathematics and physics, now machine intelligence.

B.S. in Math & Physics (Tufts) → M.S. in Analytics with a Responsible-AI concentration (Northeastern).

[1]

M.S. in Analytics — Applied Machine Intelligence.

Northeastern University · Responsible AI concentration.

Boston, MA · expected 2026

[2]

B.S. in Mathematics & Physics.

Tufts University.

Medford, MA · 2019

[3]

CareEscapes: a real-time travel-cost dashboard for medical tourism.

Team project — Finn Pounds et al., ALY6080 Integrated Experiential Learning.

Northeastern University · 2025

Code ↗

Selected graduate coursework

EAI 6400 Data Governance & Responsible AI
EAI 6000 Fundamentals of Artificial Intelligence
EAI 6010 Applications of Artificial Intelligence
EAI 6020 AI System Technologies
ALY 6015 Intermediate Analytics
ALY 6040 Data Mining Applications
ALY 6110 Data Management & Big Data

03 / Kendama

Judging, sponsorship, and a few clips.

Freestyle judge and sponsored player, mostly around Boston and the U.S. contest circuit.

Judging

FISE — International Freestyle Judge

The American judge, with Tamotsu (Gloken) & Teva (One Kendama)

European Kendama Championships

Judge — prelims (Leiden, Jan) & finals (Utrecht, May) · 2026

Battle at the Border 2026

Freestyle judge · Nashville, TN

Recap ↗

NAKO — North American Kendama Open

Minneapolis · Judge, with Christian Fraser & Kota

…and numerous smaller events.

Sponsorship

Kendama Depot

Community

mass.kendama

Local jams · Boston

Helping grow the Massachusetts kendama scene. @mass.kendama ↗

@finnpounds

Instagram

Follow along ↗

Just a Toy YouTube Cold Hands 1 Kendama Depot Cold Hands 2 Kendama Depot · TouchDesigner Cold Hands 3 Kendama Depot I Think I'm Starting to Understand 18pdox18 · effects built in TouchDesigner Onwards YouTube

04 / Bikes

Built, ridden, repaired.

Avid builder and shop-trained repair tech — every one of these is a custom build, frame up. (Two wheels are just circles, after all.)

Green vintage Fuji Sundance MTB by a lake — Fuji Sundance

Seafoam All-City Space Horse at a monument — All-City Space Horse

Oxblood All-City Gorilla Monsoon gravel bike by a lake — All-City Gorilla Monsoon

Yellow Cannondale CAAD5 R600 road bike — Cannondale CAAD5 R600

White 3x1 commuter build with red rims — 3×1 commuter

05 / About

A short note, and one on the artwork.

Finn here — a Cambridge-based builder with a B.S. in Math & Physics from Tufts, now doing an M.S. in Analytics with a Responsible-AI concentration at Northeastern. I like taking messy, real-world data and turning it into tools that help people make a decision, usually with a language model in the loop.

Away from the keyboard I play and judge competitive kendama — including the freestyle panels at Battle at the Border and the European Kendama Championships — and ride for Kendama Depot and Paradox, a brand I co-founded.

If you're hiring around applied ML, data tooling, or LLM products — or you just want to talk kendama — say hello.