Aman Kumar

AI Researcher

Mumbai, Maharashtra, India4 yrs 5 mos experience

Most Likely To SwitchAI Enabled

Key Highlights

Developed AI solutions driving significant revenue growth.
Expertise in Generative AI and Large Language Models.
Led major technical events engaging thousands of participants.

Stackforce AI infers this person is a skilled AI Engineer specializing in Generative AI and Computer Vision.

Contact

Skills

Core Skills

Generative AiLarge Language Models (llm)Computer VisionData Science

Other Skills

APIArtificial Intelligence (AI)C++Data AnalysisData MiningDeep LearningDeep Neural NetworksDeepFaceLaTeXLangChainLeadershipMATLABManagementMediaPipeOpenCV

About

As an AI Engineer with 2.8+ years of experience, I specialize in developing and deploying impactful AI solutions, particularly in Generative AI, Large Language Models (LLMs), and Computer Vision. • At Miko, I was instrumental in building VQA Systems, Multilingual Profanity Management system in autonomous robots and driving significant revenue growth ($X M+ combined) through advanced AI systems: * Pioneered an AI-driven VQA system, integrating and optimizing diverse LLMs for real-time, personalized image-based interactions, boosting Miko's revenue by $0.2X M. * Engineered an AI-driven Multilingual Profanity System, automating culturally sensitive translations across 12+ languages, generating $0.8X M by enabling global market expansion. * Developed real-time Multi-face Emotion Detection using Mediapipe and OpenCV, optimizing detection time by 50%. • My expertise spans Python , LangChain , LLM , GenAI , RAG , TensorFlow, PyTorch, Scikit-learn, and cloud platforms (GCP, AWS). I thrive on applying data-driven insights to solve complex product challenges and create innovative business solutions. Open to exciting Data Science, Machine Learning, and AI Engineering roles where I can make a genuine difference.

Experience

Miko

2 roles

AI Engineer

Jul 2023 – Present · 2 yrs 8 mos · Mumbai, Maharashtra, India

1. Developed an AI-driven VQA system, integrating and optimizing diverse LLMs for real-time, personalized image-based interactions. This system will boost Miko’s revenue by significantly enhancing customer engagement and driving global expansion.
2. Engineered an AI-driven Multilingual Profanity System, automating culturally sensitive translations across 12+ languages. This system will drive Millions in reveue for Miko, critically enabling global market entry into regions like Europe (Spain, Germany) and Asia (China, Japan).
Leading implementation of LangChain and RAG architectures for enhancing conversational AI in autonomous robots.
Actively developing Generative AI Vision Models for advanced robotic perception.

PythonLangChainRAGGenerative AILarge Language Models (LLM)Computer Vision

AI Engineering intern

Jan 2023 – Jun 2023 · 5 mos · Mumbai, Maharashtra, India

Built real-time Multi-face Emotion Detection using MediaPipe, HSE library, and OpenCV, optimized face detection time by 50% through DeepFace integration.

MediaPipeOpenCVDeepFaceComputer Vision

Science and technology @ iiit nr

Vice President

Jan 2022 – Dec 2022 · 11 mos · Raipur, Chhattisgarh, India

Led the Science and Technology club as Vice President, managing a team of 25 members.
Orchestrated and oversaw 5+ major technical events, including Hack-a-sol, DesCode 4.0, and a Bi-Monthly Hackathon, engaging 3600+ participants and coordinating with 20+ speakers.

E-cell, iiit-naya raipur

Co-Founder

Aug 2021 – Dec 2022 · 1 yr 4 mos · Raipur, Chhattisgarh, India

Co-Founded and served as Head of Events for the Entrepreneurship Cell, leading a team of 14 members.
Successfully organized the annual Entrepreneurship Summit, coordinating with 20+ speakers and attracting 2600+ participants nationwide.

Samsung india

Data Science Intern

Aug 2021 – Dec 2021 · 4 mos · Hybrid

Spearheaded the development and launch of "Collage Engine" for the Samsung Smart Call app, a pilot project impacting 748 million users.
Engineered a feature that extracts caller photos, searches galleries for common pictures, and enables real-time sharing during calls.
Optimized feature response time by over 98% (from 3 seconds to 50 milliseconds) by implementing live image scanning during video calls.
Tech Stack: Python, Deep Neural Networks, API.

PythonDeep Neural NetworksAPIData Science