Ready to innovate! 0%

Karan Patel

Master in Computer Science

Software Engineer & AI Specialist

Engineering AI that actually works—fast, reliable, ready for the real world.

About

Available for work 2+ Years of Experience
"You know, I'm
something of a scientist myself."
Memoji avatar
Available to work
2 years of experience

A Computer Science master’s student with 2+ years of hands-on engineering experience, I’ve shipped fintech products, solved real-world business problems, and launched my own tech venture. My track record proves I can design, build, and scale robust systems—delivering fresh ideas and real value from day one.

I’m happiest building things people actually use: code review tools powered by GPT, food-delivery apps loved by customers—even AI that spots disease in medical scans or environmental hazards from satellite data. Whether I’m solving a tricky backend challenge or pushing the latest in deep learning, I focus on practical, scalable solutions that make a difference—for teams, businesses, and everyday users. If you want someone who blends technical depth with real-world hustle, let’s connect.

Profile photo

Projects

Multimodal-LLM Trading Platform Ongoing

In today’s rapidly moving financial markets flooded with real-time data, this project introduces a multi-agent S&P 500 trading platform powered by specialized AI assistants with Large Language Models. Each agent analyzes distinct data streams including news, technical charts, fundamentals, sentiment, and risk. These diverse perspectives debate and combine insights to make transparent, adaptive, and robust trading decisions in real-time.

Starting with a single high-volume stock, the platform methodically scales to deliver outperformance versus traditional single-model approaches. It emphasizes resilience, explainability, and continuous learning through a stepwise roadmap with detailed reasoning reports behind every trade.

CodeKraft — AI-Powered Code Assistant

Built a VS Code extension powered by optimized GPT models. Developed scalable FastAPI backend on AWS EKS, secured with IAM, and maintained high performance with robust integration testing—making coding faster, safer, and more efficient.

PyTorchHuggingFaceGPTFastAPIAWS
GitHub

MindMate — Mental Health Companion

Chatbot integrating GPT and sentiment analysis to detect emotional tone in real time. Built in React, Firebase, AWS Lambda. Partnered with clinicians to ensure ethical, safe responses. Protected PII with Firebase security rules and at-rest encryption.

GPTSentiment AnalysisReactFirebaseAWS Lambda
GitHub

Stock Market Forecasting System

LSTM-driven prediction engine with sentiment analysis, delivering 30-day trend forecasts and real-time risk scoring. Combines Python, LSTM networks, financial APIs, and interactive dashboards for actionable market insights.

PythonLSTMFinanceData AnalysisDashboards
GitHub

Pneumonia Detection via X-ray Imaging

Built a CNN in Python and TensorFlow to classify chest X-rays for pneumonia. Used advanced image preprocessing for higher accuracy—achieving over 90%—and deployed an optimized, lightweight prototype for efficient real-world use.

TensorFlowCNNPythonMedical Imaging
GitHub

Oil Spill Classification using SAR Images

Applied advanced image processing—including median filtering, histogram equalization, thresholding, edge detection, and morphological operations—to analyze SAR satellite images for oil spill identification. Successfully segmented and classified spills across varying image conditions, aiding in environmental monitoring and remediation.

Image ProcessingSAR ImagingPython
GitHub

HungerHunt (Full Stack App)

Developed a full-stack food delivery app with React, Node.js, MongoDB, and Redis—letting users browse restaurants, place orders, and track deliveries in real time. Integrated Stripe for payments and Socket.io for instant updates. Added Redis caching to boost system performance by 20% and cut server response time by 15%.

ReactNode.jsMongoDBRedisStripeJavaScript
GitHub

Scalable Blog App Deployment Pipeline

Built a CI/CD pipeline with Jenkins, Nexus, SonarQube, and Trivy to automate build, test, and deployment of a scalable blog app on AWS EKS. Used Docker and Kubernetes for zero-downtime deployments and efficient scaling. Integrated Prometheus and Grafana for monitoring, reducing system downtime by 30% and improving performance.

AWSKubernetesJenkinsDockerPrometheusGrafanaCI/CD
GitHub

DC Real Estate & Crime Insights

Analyzed how neighborhood crime rates and socioeconomic factors impact residential property values in Washington, DC from 2014–2018. Used regression and classification to reveal key drivers in the city’s housing market.

PythonData MiningExploratory Analysis
GitHub

CineMatch

Built a personalized movie recommendation system using collaborative filtering and content-based algorithms. The app analyzes user preferences and movie metadata to suggest similar titles, enhancing user engagement and discovery.

PythonPandasScikit-learnStreamlitTMDb API
GitHub

Co-Founder

Twods Capital

At TWODS CAPITAL, we are redefining finance with a bold, out-of-the-box approach. Over the past years, we have hosted educational webinars and provided free resources to empower the younger generation with financial literacy and fintech knowledge. Our active community channel connects us directly with people, fostering responsible investing and enabling smart, confident decisions. We are passionate about leveraging data-driven, ethical strategies to create lasting impact.

Partner with us
TWODS CAPITAL

Skills & Tech

Python (expert)BashGitLinuxPyTorch TensorFlowKerasTransformersKnowledge Distillation QuantizationONNXTensorFlow LiteHuggingFaceGPT CodeBERTRAGSynthetic DataAWSDocker KubernetesFastAPIFirebaseMLflow (learning) Azure (beginner)Data PreprocessingUnit & Integration Testing Batch/Real-Time InferenceCI/CDAgile/ScrumAPI Design PandasNumPySciPyScikit-learnMatplotlib SeabornJupyter NotebookIPythonData Analysis RESTful API DesignSystem Monitoring & Observability Natural Language Processing (NLP)Data VisualizationSecurity & Compliance Technical InterviewingMentorshipProduct Management Debugging & ProfilingOpen SourceCloud Security Performance OptimizationEffective CommunicationMicroservices Technical DocumentationResponsible AICross-functional Leadership

Experience

July 2025 – Present
Technical Support Assistant
George Washington University LAW
Supporting the Law School’s technology infrastructure, I play a key role in ensuring events, classes, and presentations run without a hitch through managing multimedia systems and real-time troubleshooting.
  • Manage multimedia systems and proactively solve technical issues to prevent disruptions.
  • Deliver high-availability services and maintain robust system performance with rapid response to unexpected challenges.
  • Collaborate closely with faculty, students, and IT teams to refine cross-functional communication and user-focused problem-solving.
  • Gain hands-on experience with backend systems, technical documentation, and process automation.
  • Develop skills essential to both software engineering and AI/ML deployment through practical, real-world application.
Feb 2025 – Apr 2025
Research Assistant
George Washington University SEAS
At GWU as a Research Assistant, I designed and developed a scalable AI-powered code checking and exam testing system that leverages large language models—including GPT, CodeBERT, and CodeT5—to automatically review student code submissions in real time.
  • Built Python-based REST APIs and automated the grading pipeline with CI/CD.
  • Deployed the service on the cloud using Docker and Kubernetes, ensuring high reliability and minimal downtime.
  • Achieved over 95% error-detection accuracy through rigorous unit and integration testing.
  • Monitored system performance and iterated features based on student and faculty feedback.
  • Gained hands-on experience in full-stack engineering, applied machine learning, and cloud infrastructure.
  • Delivered a powerful, user-friendly system that enhances teaching and learning outcomes.
Jun 2023 – Jul 2024
SDE 1 & Business Solution Analyst
VE, Mumbai, India
I spent a year at VE as both a Software Development Engineer and a Business Solution Analyst, leading the design, development, and deployment of cloud-based web and mobile platforms with Node.js, MongoDB, Docker, and Kubernetes.
  • Delivered REST APIs, microservices, and robust CI/CD pipelines—driving feature launches, automated testing, and reliable incident response.
  • Worked closely with product and ops teams, balancing engineering rigor with business insight.
  • Leveraged ML/AI to build predictive analytics for inventory, NLP chatbots to automate support, and dashboards to convert raw data into actionable decisions.
  • Led user interviews, gathered feedback, and mentored junior engineers—ensuring clean, scalable, and valuable code.
  • Accelerated critical workflows, delivered intelligent product features, and built a codebase and culture prepared for future growth.
Sep 2019 – Present
Co-Founder
TWODS CAPITAL
Built algorithmic trading systems in Python, implementing version-control workflows (Git) and continuous testing practices to ensure code quality and reliability.
  • Hosted seminars on trading algorithms and technical analysis, emphasizing responsible and confidentiality-aware data handling.
  • Developed open-source tools and educational content that significantly expanded a community of retail traders.
  • Contributed to a 50% increase in member engagement and knowledge sharing through community building efforts.

Education

Jul 2024 – May 2026
Master of Science, Computer Science
The George Washington University
Coursework: Distributed Systems, Advanced Software Engineering, Deep Learning
Jul 2018 – May 2024
Bachelor of Technology, Computer Engineering
NMIMS, Mumbai, India
Honours in AI; Coursework: Machine Learning, Artificial Intelligence

Get in Touch