Senior VLM Developer – Vision Language Models

Technozis · Bangalore, India

Senior VLM Developer – Vision Language Models

Location: Bangalore, India

Experience: 7.5+ Years

Joining: Immediate – September 2026

We are looking for an experienced Senior VLM Developer with strong foundations in Machine Learning and Deep Learning to build, fine-tune, and evaluate Vision-Language Models (VLMs) for multimodal perception and understanding applications.

The role involves developing large-scale training pipelines, curating multimodal datasets, experimenting with model architectures, and optimizing models for downstream applications.

Key Responsibilities

• Design, implement, and fine-tune VLM architectures such as LLaVA, Qwen-VL, or similar models .
• Develop data preprocessing and training pipelines for large-scale multimodal datasets.
• Optimize model performance across visual-language benchmarks and internal use cases.
• Collaborate with research and synthetic data teams to integrate generated data into model training.
• Conduct ablation studies and experiments to evaluate model performance.
• Maintain documentation of experiments, model behavior, and evaluation results.
• Develop and optimize ML/DL solutions for complex business problems.
• Work independently on statistical, machine learning, and research-oriented projects.
• Collaborate with cross-functional teams to translate research and experimentation into practical AI solutions.

Required Skills

• 7.5+ years of experience in Machine Learning / Deep Learning / Data Science.
• Strong hands-on experience with Vision-Language Models (VLMs) and multimodal AI.
• Experience designing and fine-tuning VLM architectures such as LLaVA, Qwen-VL, or equivalent models .
• Strong hands-on expertise in PyTorch .
• Experience with PyTorch Lightning .
• Strong experience with the Hugging Face ecosystem.
• Strong understanding of Transformers and vision-language architectures .
• Experience with multimodal fusion techniques.
• Experience with distributed training frameworks .
• Hands-on experience with large-scale model training, fine-tuning, and evaluation.
• Experience with PEFT / parameter-efficient fine-tuning and/or reinforcement learning techniques.
• Strong experience in data preprocessing and multimodal dataset preparation .
• Experience working with synthetic data is preferred.

Mandatory Domain Experience

Candidates must have experience in at least one of the following domains:

CPG | Retail | Pharma

Candidates without relevant CPG, Retail, or Pharma domain experience should not be considered.

Education

B.Tech / M.Tech / Ph.D. in:

• Computer Science
• Artificial Intelligence
• Machine Learning
• Data Science
• Or a related technical discipline

Joining Requirement: Immediate joiners only.

Candidates must be able to join on or before 30 September 2026 .

Candidates with a joining date after 30 September 2026 will not be considered.