Prakash Kumar

About Prakash Kumar

I\'m a Machine Learning and Backend Systems Engineer with hands-on experience in building real-time APIs, scalable microservices, and intelligent data pipelines. Currently pursuing a B.Tech in Computer Science (AI & ML specialization) at Lovely Professional University, with a CGPA of 8.3/10 and 250+ DSA problems solved. During my recent internships at ByteBix Technologies and CodeSpaze, I designed and deployed:• A multilingual caption generation system using OpenAI Whisper + SeamlessM4T — 92% transcription accuracy, 40% latency reduction • A real-time ride-hailing backend using FastAPI + graph-based matching — improved matching accuracy by 88%, reduced ETA by 23% • Predictive models for churn, pricing, and fraud detection — achieved 94%+ model accuracy and recall through advanced ML tuning I’ve also explored financial AI, including sentiment-based equity movement prediction using TF-IDF and news datasets (87% accuracy on Nifty50-classified stocks). Skills & Tools:Python, SQL, C++, JavaScript, Swift, React.js, ASP.NET MVC, Scikit-learn, TensorFlow, Keras, FastAPI, Docker, Power BI, PostgreSQL, AWS, Git, Jupyter, Matplotlib, Seaborn Passionate about combining ML, data science, and backend development to build meaningful, user-focused solutions. Open to full-time roles or internships in ML Engineering, AI R&D, FinTech, or Backend Development.Let’s connect if you\'re building something exciting!

Experience

1. Machine Learning Intern

   Bytebix Technologies

   Jan 2025 — Present

   Developed a multilingual captioning system using OpenAI Whisper and SeamlessM4T APIs, achieving over 92% transcription accuracy across 10+ languages and reducing latency by 40% through asynchronous FastAPI microservices.• Built and deployed a backend ride-matching system using Google Maps API and real-time scoring logic; improved matching accuracy by 88% and decreased estimated arrival time by 23% through bipartite graph matching (Hungarian algorithm).• Worked cross-functionally with frontend and devops team members to containerize and test ML microservices using Docker and Postman.• Contributed to scalable backend infrastructure design, emphasizing modularity and reusability in production ML systems.• Technologies used: Python, FastAPI, Scikit-learn, Docker, Whisper API, SeamlessM4T, Google Maps API, Git, Pandas, NumPy.

Education

• R.L.S.Y Inter College Muzaffarpur

  High School Diploma, Science

  2016 — 2019

• Lovely Professional University

  Bachelor of Technology - BTech, CSE (Hons.)

  2020 — 2024

• Lovely Professional University

  Bachelor of Engineering - BE, Computer Science

• Lovely Professional University

  Bachelor of Technology, Computational Science

  2020 — 2024

• Lovely Professional University

  Bachelor of Technology - BTech, Computer Science