Open to opportunities

hardik juneja

@hardikjuneja

AI Engineer at APML developing face recognition, multimodal AI, and voice-call processing pipelines with local language models.

India

Message

At APML, I develop face recognition systems using deep learning embeddings, and multimodal solutions for image understanding, OCR post-processing, and structured data extraction.

I also developed an end-to-end voice-call processing pipeline for transcription, Hinglish speech recognition, and extracting structured household articles. Previously, I built Python chat pipelines and a RAG chatbot at Laiqa Wellness Private Limited, and developed Django APIs and contributed to an Airflow workflow at ITECH MISSION PVT. LTD.

Experience

Work history, roles, and key accomplishments

Current

AI Engineer

Current

APML

Nov 2025 - Present (11 months)

Developed face recognition systems using deep learning-based embeddings and built multimodal AI solutions with Vision-Language Models and local LLMs. Deployed and optimized local AI inference for scalable production voice-call processing.

Deep Learning Face Recognition Vision Language Models LLM OCR Whisper Qwen ASR Gemma Qwen fastAPI VLLM Celery Redis Audio Transcription Hinglish Speech Recognition

Python Developer

Laiqa Wellness Private Limited

Jun 2025 - Nov 2025 (5 months)

Automated export and storage of WhatsApp chats using a Python-based Playwright pipeline and implemented a RAG pipeline for a context-aware chatbot. Fine-tuned a large language model on domain-specific data to improve query handling.

Python Playwright RAG Chatbot Development Fine Tuning JSON Large Language Models

Solution Python Developer

ITECH MISSION PVT. LTD

May 2022 - May 2025 (3 years)

Developed REST APIs using Django and MongoDB and contributed to an automated workflow with Airflow for data ingestion and processing. Dockerized the application and assisted in deployment.

Python Django MongoDB REST APIs Airflow Docker RabbitMQ Celery