AI-ML Engineer

AI-ML Engineer
Location: HYD, India

Role Overview:
We're seeking an AI/ML Engineer to explore, evaluate, and optimize open-source models
for multimodal classification and annotation tasks. You'll work across the full lifecycle
from model exploration and fine-tuning to optimizing inference pipelines and integrating
with Meeami's data platform.
This role is ideal for someone passionate about applied machine learning eager to
experiment with emerging architectures, improve model performance, and bring AI
capabilities into production workflows.

Key Responsibilities:
Research, evaluate, and benchmark open-source models for classification,
detection, and annotation tasks across text, audio, image, and video.
Fine-tune and optimize models for accuracy, latency, and resource efficiency on
real-world datasets.
Build and maintain inference pipelines and APIs (using FastAPI, TorchServe, or
Triton) for seamless integration with backend services.
Collaborate with backend and data teams to design data flows for training, testing,
and evaluation.
Perform exploratory analysis and visualization to understand dataset quality and
model behavior.
Define and track evaluation metrics to continuously measure model performance
and reliability.

Contribute to early training pipelines, experiment tracking, and data versioning
initiatives.

Qualifications:
36 years of experience in applied machine learning or AI engineering.
Strong programming skills in Python, with hands-on experience in PyTorch or
TensorFlow.
Familiarity with data preprocessing and augmentation for text, audio, image, or
video datasets.
Experience running and profiling models for inference (GPU/CPU) using ONNX,
TorchScript, or TensorRT.
Working knowledge of FastAPI or similar frameworks for serving ML models.
Practical understanding of Git, CI/CD, Docker, and Linux environments.
Comfort working in cloud environments (AWS/GCP/Azure) and collaborating in
agile, cross-functional teams.

Preferred Experience:
Experience with audio processing, speech recognition, or computer vision models
(classification, segmentation, detection).
Familiarity with annotation workflows and dataset QA.
Understanding of model evaluation metrics (precision, recall, F1, mAP, AUC).
Exposure to model optimization techniques (quantization, pruning, distillation).
Experience with ML experiment tracking and dataset versioning tools (MLflow, DVC,
Weights & Biases).
Bonus: Knowledge of transformer-based and multimodal architectures (Whisper,
CLIP, CLAP, LLaVA, SAM, etc.).
.