Manas Jha

Machine Learning Engineer | AI/ML Specialist
Delhi, IN.

About

Highly motivated Machine Learning Engineer and Co-Founder with proven expertise in developing and deploying advanced AI/ML solutions, including LLMs and computer vision systems. Successfully leveraged Python and Deep Learning to build high-accuracy classification systems and real-time inventory solutions, notably reducing losses by 10% and improving response times by 30% for over 2000 users. Seeking to apply strong technical skills and leadership capabilities to drive innovative, impactful solutions in a dynamic engineering environment.

Work

Sarthi.me
|

Co Founder

Noida, Uttar Pradesh, India

Summary

Led AI and backend R&D for an early-stage startup, developing innovative systems to enhance user engagement and optimize project execution.

Highlights

Directed AI and backend R&D initiatives, developing Deep Learning and FastAPI systems to interpret and respond to human emotions, significantly enhancing user engagement in natural interactions.

Supervised and mentored a team of interns, overseeing research tasks, reviewing work, and steering prototype development, which improved project efficiency and fostered intern skill development.

Marshee Pet Tech
|

Machine Learning Engineer Intern

Gurugram, Haryana, India

Summary

Designed and deployed a high-accuracy dog breed classification system, ensuring robust performance and enhancing user experience.

Highlights

Designed and deployed a dog breed classification system with 88% accuracy, enhancing user experience by integrating robust unit and integration testing to ensure system reliability.

Agro Farm Ventures Pvt Ltd
|

Computer Vision Engineer Intern

Noida, Uttar Pradesh, India

Summary

Developed a real-time, ML-powered camera-based inventory system, leveraging automated test case design to optimize stock management.

Highlights

Developed a real-time, ML-powered camera-based inventory system leveraging automated test case design to verify early stock-out predictions, which reduced losses by 10% and boosted sales.

Samsung Innovation Campus
|

Trainee

Noida, Uttar Pradesh, India

Summary

Completed intensive training in machine learning and deep learning, developing and validating an AI agent for dietary recommendations.

Highlights

Completed comprehensive training in machine learning, neural networks, and deep learning, applying automated testing principles and quality checks.

Developed and validated a Gym Nutrition AI Agent, enhancing the accuracy of dietary recommendations for users through rigorous testing and quality assurance.

Education

Gautam Buddha University
Greater Noida, Uttar Pradesh, India

B.Tech

Information Technology

Grade: 8.0/10 CGPA

Courses

Machine Learning

Data Science

Cloud Computing

Awards

Winner: Agritech Hackathon (Microsoft Azure Sponsored)

Awarded By

Microsoft Azure

Developed an AI-powered stubble fire detection system using satellite imagery, enabling real-time monitoring and early alerts. Designed an interactive dashboard for fire event tracking, showcased to Microsoft engineers during the Grand Finale.

Skills

Programming Languages

Python, Java.

Backend Development

FastAPI, Flask, Git, JWT.

Databases

SQL, MongoDB, Neo4j, Supabase, Vector Databases, Knowledge Graph.

AI & Machine Learning

Machine Learning, Generative AI, LLMs, LangChain, Computer Vision, Deep Learning, Natural Language Processing, Neural Networks.

Cloud & DevOps

Google Cloud (GCP), Docker, CICD.

Automation

Selenium, API Integration.

Libraries & Frameworks

scikit-learn, TensorFlow, PyTorch, NumPy, Pandas, Matplotlib, Seaborn.

Quality Assurance

Software Testing Concepts, Mobile App Testing.

Methodologies

Agile Methodologies.

Projects

Stubble Fire Detection & Prevention System

Summary

Designed a prototype for stubble burning detection using satellite imagery and custom machine learning models to accurately identify fire hotspots in agricultural areas.

RAG-Based AI Chatbot for College Queries

Summary

Developed a Retrieval-Augmented Generation (RAG) chatbot using a vector database and knowledge graph DB to provide accurate answers to college queries with high contextual accuracy.