Deven Patel — Engineer & Builder

About me

Builder by instinct,
engineer by craft.

I'm a software engineer and AI builder with a deep background spanning machine learning research, founding-stage startups, and production-scale systems. My work lives at the intersection of AI, cloud infrastructure, and user-centric product design.

Most recently I'm driving a global healthcare interoperability platform at Kai Nest, building around openEHR/FHIR standards across five countries. Before that, I was a Founding Engineer at Mem0 (formerly Embedchain), where I helped build an open-source RAG framework that's gathered over 40,000 stars on GitHub.

I hold a Master of Engineering in Computer Science from Penn State, where I also did research at the Deep Learning Lab on landslide prediction and at the Undergraduate Deep Learning Lab on telescope image analysis.

Outside of work I recharge on the football pitch, cycling trails, and in books. I believe movement and curiosity are the best complements to a life in tech.

Let's talk →

Current Focus

🏥 Healthcare AI & Interoperability (openEHR / FHIR) 🤖 RAG & LLM systems ☁️ Cloud-native architecture (GCP / AWS)

Interests

Football ⚽
Cycling 🚴
Hiking 🧦
Reading 📚
Open Source 🛠

Looking for

Full-time roles in AI/ML Engineering, Full-Stack Development, or Cloud/Platform Engineering.

Work experience

Where I've built.

              Present
              Dec 2024
            
~3 mos
Kai Nest Pvt Ltd
Project Manager
🇮🇳 Gujarat, IN
Defined and drove execution of a healthcare platform with openEHR/FHIR interoperability across India, USA, Canada, UK, and France.
Led refinement of 10+ core workflows including multi-modal chatbot-driven health history capture, goal setting, resource discovery, and clinician-validated AI recommendations.
Directed modular architecture featuring Rule Author, Prompt Author, AI Interface, and EHR knowledge graph for personalized wellness/clinical outputs.
Established global security standards: encryption, audit trails, multi-lingual support (EN/HI/FR/GU/ES), and multi-modal inputs (text, audio, imaging, ECG).

              openEHRFHIRAIAgileHealthcare
            

Jan 2024 Oct 2023

4 mos

Mem0 (pka Embedchain)

Founding Engineer

🇺🇸 Mountain View, CA

Built an open-source Retrieval Augmented Generation (RAG) framework for developing custom GPTs on unstructured data.
Framework received community acclaim — 40k+ GitHub stars on mem0ai/mem0.

PythonLLMRAGAIOpen Source

Jun 2023 Jul 2021

2 yrs

Youpi Inc.

Founding Engineer

🇺🇸 Mountain View, CA

Managed a team of engineers and product designers, ensuring high-quality daily releases in a fast-paced environment.
Developed generative AI features to create personalized portrait images from user-provided creative prompts.
Reduced video serving size by 80% via video transcoding in the backend.
Designed and implemented backend infrastructure on Google Cloud; built camera-based social media app in Swift.

SwiftPythonNodeJSGoogle CloudFirebaseGenAI

May 2021 Aug 2019

1 yr 9 mos

Deep Learning Lab, Penn State

Research Assistant

🇺🇸 University Park, PA

Collaborated with Civil Engineers on landslide prediction using remotely sensed images.
Applied data augmentations to reduce class imbalance, improving performance by 5%.
Incorporated DEM information into the encoder to boost accuracy by 10%, reaching 92%.

TensorFlowPythonGoogle CloudDeep Learning

May 2019 Jun 2018

1 yr

UDLL, Penn State

ML Research Assistant

🇺🇸 State College, PA

Built a deep learning model detecting alpha-particle defects in radio telescope images — 78% accuracy on a highly imbalanced dataset (60% images with 0-1 pixel defects).

TensorFlowPythonComputer Vision

Dec 2019 Aug 2017

2 yrs 4 mos

D.A.T.A. Labs, Penn State

Undergraduate Research Assistant

🇺🇸 State College, PA

Improved a Child Safety Game for low-income parents.
Developed an end-to-end website for users to capture vitals and download vital data (videovitals.org).
Filtered data to train a deep neural network generating 3D objects from 2D projections.

UnityDjangoReactJSAWS

Aug 2017 Jun 2017

3 mos

KOGENTiX Inc. (acq. Accenture)

Data Engineer Intern

🇺🇸 Schaumburg, IL

Built distributed applications to solve large-scale data processing problems using MapReduce, significantly reducing processing time.
Used HQL and SparkSQL for data analysis; optimized cloud deployments using modern big data tech.

HadoopMapReduceHQLLinux

Selected projects

Things I've built.

🧠

40k ★

↗

Mem0 (RAG Framework)

Universal memory layer for AI agents. An open-source RAG framework that lets you build custom GPTs on unstructured data — one of the most-starred AI libraries on GitHub.

PythonLLMRAGOpen Source

🗺️

BuddyMap

Social app to locate nearby friends and discover local events. Includes real-time location sharing, messaging, group creation, and event management.

DjangoReact NativePython

🌋

LandslideNet

Deep learning model for landslide detection using remotely sensed images, collaborating with Civil Engineers. DEM data integration improved accuracy to 92%.

TensorFlowPythonGoogle Cloud

📂

Peer-to-Peer File System

Torrent-like P2P file distribution in C. Multithreading and piece-selection heuristics for minimal distribution time; fault-tolerant with crash recovery.

CLinuxTCP SocketsMultithreading

🗃️

Parallel File System (PFS)

Multi-client PFS with files striped across servers. Includes client-side invalidation cache and Distributed Metadata Manager — built with gRPC.

C++gRPCLinuxCaching

🔤

Part-of-Speech Tagger

Hidden Markov Model for POS tagging trained on the Brown corpus. Achieved 94% accuracy through careful transition and emission probability optimization.

PythonNLPHMM

Deven
Patel.

Builder by instinct,
engineer by craft.

Where I've built.

What I work with.

Things I've built.

Where I learned.

Let's build
something great.

DevenPatel.

Builder by instinct,engineer by craft.

Where I've built.

What I work with.

Things I've built.

Where I learned.

Let's buildsomething great.

Deven
Patel.

Builder by instinct,
engineer by craft.

Let's build
something great.