Building AGI · TDK SensEI · Pittsburgh

Pranav
Kanuparthi

My mission is building AGI — intelligent systems that don't just predict, but understand, act, and collaborate with humans to solve the world's hardest problems. Right now: agentic AI and industrial intelligence at TDK SensEI, where sensor signals become autonomous decisions at the edge.

Agentic AI Industrial ML Edge AI Sensing Systems AWS Bedrock LangChain · LangGraph
5+
Patents
$20M
Downtime Saved
98%
Pose Accuracy
CES
Innovation Award
00

What I Bring

I'm at my best when the problem is half-defined, the stakes are real, and nobody's built it before. From co-founding a CES Award-winning hardware-AI startup to leading multi-agent AI systems at TDK SensEI, the thread is always the same: figure it out fast, build something that works, then make it excellent.

🧭
Mindset
Navigates Ambiguity with Confidence

Comfortable where requirements are fuzzy and the path undefined. I've shipped products, filed patents, and run novel research — in domains where there was no playbook.

Speed
Rapid Prototyper

From idea to working demo fast — multi-agent fault detection systems, neuromorphic architectures, adversarial defenses. I build first, then iterate to rigour.

🔬
Thinking
First-Principles Engineer

When existing approaches fail, I design new ones. Hierarchical agent architectures, temporal neural networks — five patents aren't a hobby, they're a habit.

🧱
Range
Full-Stack ML Builder

Raw sensor hardware → signal processing → model architecture → cloud deployment. No handoff required between research and production.

🎯
Drive
Mission-Oriented, Not Task-Oriented

One consistent question across 9 years: how do we make machines that truly understand their environment? Hard problems. Real stakes. Always.

🤝
Leadership
Builder of Teams and Culture

Cross-functional delivery across engineering, data, QA and product. Presented to executive leadership. Mentored engineers. Momentum, not just output.

01

Skills

Knowledge Areas
Deep Learning
Machine Learning
Agentic AI Systems
Sensor Systems Dev.
Data Analytics
Frameworks & Tools
PyTorch · TensorFlow · Keras
LangChain · LangGraph
PySpark · Pandas · NumPy
Scikit-Learn · Pydantic
Cloud & Infrastructure
AWS Bedrock
Lambda · S3 · DynamoDB
ECR · Docker
Languages
Python
C / C++
SQL
02

Experience

TDK SensEI
Full-time
Sep 2021 – Present
Pittsburgh, PA
formerly Qeexo
Senior Machine Learning EngineerSep 2021 – Present
  • Architected and led development of a production-ready multi-agent AI system for industrial asset diagnostics — three-agent hierarchical architecture using AWS Bedrock, LangChain, and LangGraph for real-time fault detection in rotating equipment.
  • Eliminated approximately $20M in annual downtime losses for large manufacturing clients; improved data pipeline efficiency by 62% memory reduction and 20% latency improvement.
  • Developed automated clustering validation and gating mechanism for CBM_ML platform, eliminating manual review and improving model accuracy, deployment speed, and labeling efficiency across equipment types.
  • Drove cross-functional delivery across engineering, data, QA, and product; presented technical roadmap and architecture decisions to executive leadership.
  • Mentored two engineering interns; delivered invited talk on embedded ML systems at ARM DevSummit 2024. ↗ Watch
$20M Downtime Saved 62% Memory Reduction 20% Latency Improvement ARM DevSummit Speaker
Carnegie Mellon University
Research · 8 mos
May – Dec 2020
Pittsburgh, PA
Graduate Research AssistantNeuromorphic Computer Architecture Lab · Advisor: Prof. John Shen
  • Developed Temporal Neural Network (TNN) architectures for Keyword Spotting and time-series classification using PyTorch, targeting Neuromorphic Sensory Processing Units.
  • Achieved 73% accuracy on Spoken MNIST under strict neuromorphic power constraints.
Wellnesys Inc.
Co-Founder
Oct 2018 – Jan 2020
Bangalore, India
Co-Founder & Chief Innovation OfficerOct 2018 – Jan 2020
  • Built YogiFi — a CES Innovation Award-winning, patented AI + pressure sensor platform for real-time yoga pose classification and personalised guidance. Led a twelve-member cross-functional team.
  • Engineered CNN on pressure matrix data achieving 98% pose classification accuracy; technology underpins 3 granted patents.
03

Projects

P — 01
Signal ProcessingPulse OximetryEdge
Breathe-right

Real-time respiratory monitoring via pulse oximeter. Built for athletes and COVID-19 patients — continuous, non-invasive, edge-deployable.

P — 02
Computer VisionOpenPoseHealthcare
Guardian

Camera-based gait analysis for Huntington's disease patients. Quantifies chorea progression — turning qualitative clinical observation into trackable data.

P — 03
ML SecurityPyTorchResNet
Neural Network Defense

Three-stage adversarial defense strategy for image classifiers. +40% accuracy recovery on an attacked ResNet-32 under active attack.

P — 04
Distributed MLRecommenderPipeline
Related Rhythms

Distributed ML pipeline for music recommendation. End-to-end pipeline design — from audio feature ingestion to ranked personalised output.

04

Patents & Publications

Patents
System and non-intrusive method for exercise posture detection
GrantedUS 20200254299 A1 · IN 201941005350
System and method to monitor an exercise posture of at least one user
GrantedIN 201841006784 · WO 2019162743A1
A flexible substrate and a method thereof
GrantedIN 201841006773 · WO 2019162742A1
System and method for performance evaluation of an exercise performed by a user
PendingIN 202041000701
Real-time offside offense determination system and a method thereof
PendingIN 201841046250
Talks & Publications
Embedded ML for Industrial IoT Applications
ARM DevSummit 2024
↗ Watch presentation
The Changing Nature of Human Adjudication
ITSMTGPD 2019 — IEEE Technology and Society Magazine
05

Education

Carnegie Mellon University
Master of Science in Electrical and Computer Engineering
Specialised in Machine Learning
May 2021 · Pittsburgh, PA
RV College of Engineering
Bachelor of Engineering in Electronics and Instrumentation Engineering
 
May 2017 · Bangalore, India
06

Credentials

CES
Innovation Award — YogiFi
Consumer Electronics Show
Udacity
Deep Learning Nanodegree
Oct 2017
DataCamp
Machine Learning with Python
Sep 2017
DataCamp
Data Manipulation with Python
Sep 2017
DataCamp
Importing & Cleaning Data
Sep 2017
US Sailing
Small Boat Sailor
Jul 2025
07

Beyond the Terminal

Add your sailing photo
On the water
Certified Sailor
US Sailing certified. Reading wind and water scratches the same itch as debugging a model — pure systems thinking, no safety net.
🏎️ Add your motorsport photo
Speed & precision
Motorsport
Machines at the edge of their limits. The engineering is half the attraction.
🎯 Add your clays photo
Focus & timing
Sporting Clays
Every target is a different problem. Fast pattern recognition, total presence — no room for distraction.
🥊 Add your boxing photo
Physical discipline
Boxing
Footwork, timing, controlled aggression. The sport that rewards preparation more than almost any other.
🎮 Add your gaming photo
Strategy & reflex
Gaming
Competitive at heart. Pattern recognition, fast decisions under pressure — good practice for production ML.
📚 Add your photo
Always learning
Reading
Across tech, history, and whatever's next. Curiosity isn't a trait — it's a practice.
The best engineers aren't defined
by their stack.
They're defined by their range.

On the water, on the range, trackside, or in a game — the same instincts show up everywhere: read the situation fast, commit fully, and execute cleanly. A full life makes for a sharper engineer.

Open to the right problem

Let's build
something
10x.

I'm looking for senior ML / agentic AI roles where the problem is hard, the stakes are real, and the solution doesn't exist yet. If that's what you're building — let's talk.

pranavkanuparthi@gmail.com