Appu K

AI Researcher

Thrissur, Kerala, India1 yr 3 mos experience

AI EnabledAI ML Practitioner

Key Highlights

Led development of a Fortune 500 retailer's AI system.
Designed innovative virtual try-on systems for e-commerce.
Mentored over 10 interns in AI/ML projects.

Stackforce AI infers this person is a Generative AI and Agentic AI specialist in the retail and e-commerce sectors.

Contact

Skills

Core Skills

Generative AiAgentic AiSystem Architecture

Other Skills

AutogenRetrieval-Augmented Generation (RAG)KafkaCeleryRedisLangChainLangGraphMediapipeRTMPoseONNXGemini Live APIOpenAI's Realtime APIMicrosoft TeamsMySQLJakarta Persistence

About

AI/ML Engineer working primarily on Generative AI and Agentic AI systems, with experience in designing and deploying scalable, production-grade solutions for large-scale retail and e-commerce environments. Hands-on in building multi-agent architectures, intelligent automation workflows, and real-time AI services. Has practical experience across Generative AI applications, RAG pipelines, virtual try-on systems, demand sensing, and workforce optimization, along with solid exposure to machine learning and deep learning fundamentals. Currently pursuing an MTech in Artificial Intelligence and Machine Learning at BITS Pilani (WILP), with an expected graduation in January 2028, and planning a Deep Learning specialization to deepen ML and DL expertise.

Experience

1 yr 3 mos

Total Experience

1 yr 3 mos

Average Tenure

1 yr 3 mos

Current Experience

Litmus7

2 roles

Associate Engineer - AI/ML

Feb 2025 – Present · 1 yr 3 mos · Kochi, Kerala, India · On-site

Led the development of a production-grade, event-driven Agentic AI system for automated incident detection and resolution in a Fortune 500 retailer's enterprise Order Management System, managing a 4-person cross-functional team while driving client interactions, architecture decisions, and sprint demos.
Architected real-time alert ingestion pipelines using Kafka, Celery, and Redis for asynchronous processing, enabling end-to-end automated anomaly detection across large-scale order workflows deployed on AKS in the Azure ecosystem.
Designed a multi-agent resolution system with 6+ specialized agents using LangChain, LangGraph, and MCP, with LiteLLM as a multi-provider LLM gateway. Integrated human-in-the-loop approval workflows and RAG with pgVector for contextual grounding.
Designed and built a virtual try-on system for jewelry across 3 categories (necklaces, bracelets, rings) using Mediapipe, RTMPose, and custom keypoint estimation on 2D images; trained and exported models to ONNX format for mobile AR deployment.
Delivered 3+ client-adopted AI POCs, including a real-time store assistant using Gemini Live API, a VoIP calling agent using OpenAI's Realtime API, and agentic customer support bots integrated with Microsoft Teams.
Conducted applied research on demand sensing and workforce optimization for retail BOPIS operations, synthesizing findings from academic and industry publications into strategy documents.
Mentored 10+ interns and freshers across multiple cohorts, currently guiding a team of 5 building an end-to-end Agentic AI platform.

AutogenRetrieval-Augmented Generation (RAG)KafkaCeleryRedisLangChain+9

Associate Engineer - Trainee

Jul 2024 – Feb 2025 · 7 mos · Kochi, Kerala, India · On-site

Completed intensive training in Java and backend engineering, then transitioned to the AI/ML team where I began working on Generative AI and Agentic AI projects.
Built initial POCs in multi-agent systems and RAG pipelines that led to being assigned as lead on the OMS incident analysis project.

MySQLJakarta Persistence