Summary
Overview
Work History
Education
Skills
Projects
Certification
Social
Timeline
Generic

Khushmeet

Sirsa

Summary

AI Engineer with expertise in developing and deploying production-ready AI applications using generative AI, NLP, and cloud platforms like Azure OpenAI and AWS Lambda. Skilled in prompt engineering, deep learning, and data preprocessing, with a proven ability to enhance efficiency and accuracy. Experienced in collaborating with cross-functional teams to deliver innovative AI solutions. Committed to staying updated with the latest AI/ML trends and technologies.

Overview

5
5
years of professional experience
1
1
Certification

Work History

AI Engineer

Prime Focus Technologies
Bangalore
05.2023 - Current
  • Developed production-ready Generative AI workflows for subtitling and localization, achieving nearly 100% accuracy improvement using NLP techniques like dialogue priority and speaker diarization.
  • Designed a multi-agent AI system on AWS Bedrock, evolving from chain-of-thought to multi-agent architecture, enhancing subtitling workflows for major clients.
  • Led the Title & Synopsis Agent development using Langchain, NLP, and Azure OpenAI, implementing configurable prompt optimization to increase metadata generation efficiency by 60%.
  • Performed data preprocessing tasks, including text normalization, chunking, and semantic matching, to optimize language model inputs for subtitling and localization.
  • Collaborated with content localization teams to create ground truth subtitles for word error rate analysis and with data analysts to visualize performance metrics using Power BI dashboards.
  • Enhanced sports analytics event detection using prompt engineering and Python automation, improving accuracy and efficiency.
  • Optimized multi-language speech-to-text processing through parallel processing and refined speaker tagging per industry standards.
  • Engineered an AI tool for extracting structured script data with 95% accuracy, streamlining pre-production using NLP and data pipelines.

Freelance AI & Localization Engineer

DayTranslations, Elite Asia, Tridindia, Baybel
Remote
02.2020 - 05.2023
  • Designed AI-powered workflows for subtitling, transcription, and translation, achieving 80-90% improvements in speed and accuracy for clients like Netflix and Disney+ Hotstar.
  • Applied data transformation and normalization to prepare multilingual content for AI processing, ensuring consistency across languages.
  • Developed a prompt regenerator to optimize inputs for AI image and video generation, enhancing output precision.

Education

Bachelor of Technology - Computer Science And Engineering

Maharaja Ranjit Singh Punjab Technical University
Punjab

Skills

  • Prompt Engineering: Advanced techniques including configurable prompt optimization and few-shot learning
  • Machine Learning: Knowledge of algorithms (eg, linear regression, decision trees, clustering)
  • Deep Learning: Neural networks, large language models (LLMs)
  • Natural Language Processing (NLP): Text processing, semantic analysis, language generation
  • Cloud Platforms: Azure OpenAI, AWS Lambda, AWS Bedrock
  • Programming Languages: Python
  • Tools & Frameworks: Langchain, Hugging Face, Whisper, FAISS, FastAPI, Streamlit
  • Data Visualization: Power BI
  • Databases: MongoDB, Elasticsearch
  • DevOps: Docker, Git

Projects

  • Title & Synopsis Agent: Automated video metadata generation using Langchain, NLP, and Azure OpenAI. Employed configurable prompt optimization and few-shot learning to tailor outputs to user-specified styles, achieving a 60% increase in processing speed.
  • Localization Agent: Built a multi-agent AI system on AWS Bedrock, transitioning from chain-of-thought to multi-agent workflows, integrating transcription, translation, and quality control for culturally relevant subtitles.
  • RAG Multi-Modal Chatbot: Developed a chatbot integrating PDF, audio, and video inputs using Mistral LLM, Whisper AI, Langchain, Streamlit, and FAISS, enabling context-aware responses.
  • Master Script Creation Engine: Engineered an AI tool for extracting scene-level script data with 95% accuracy, using NLP and data preprocessing pipelines for cleaning and structuring.
  • Basketball Segment Analysis System: Utilized GPT vision models and advanced NLP for event detection in basketball games, achieving over 90% accuracy in identifying key moments.

Certification

  • AI Agents in LangGraph – DeepLearning.AI
  • Understanding Prompt Engineering – Data Camp
  • ChatGPT Prompt Engineering for Developers – DeepLearning.AI
  • Prompt Engineering for Vision Models – DeepLearning.AI
  • Open Source Models with Hugging Face – DeepLearning.AI
  • Prompt Compression and Query Optimization – DeepLearning.AI
  • IELTS (Band Score: 7 | C1 Proficiency)
  • HTML, CSS, and JavaScript for Web Developers – Johns Hopkins University (Coursera)

Social

Khushmeet Arora - AI Engineer - Prime Focus Technologies | LinkedIn

Timeline

AI Engineer

Prime Focus Technologies
05.2023 - Current

Freelance AI & Localization Engineer

DayTranslations, Elite Asia, Tridindia, Baybel
02.2020 - 05.2023

Bachelor of Technology - Computer Science And Engineering

Maharaja Ranjit Singh Punjab Technical University

Similar Profiles

CREATE PROFILE
Khushmeet