🧬 Python Developer – AI Automation & Custom Model Specialist

🌟 Position Overview

Python Developer

🧠 Key Responsibilities🤖 AI Workflow Automation & Orchestration

• Build intelligent automation pipelines for

  clinical workflows

• Orchestrate real-time and batch AI workflows using

  Airflow

  ,

  Prefect

  , or

  Dagster

• Develop

  event-driven

  architectures and

  human-in-the-loop

  validation layers
• Automate data ingestion, processing, and inference pipelines for medical data

🧪 Custom AI Model Development

• Train

  custom NLP, CV, or multi-modal models

  for medical tasks
• Fine-tune

  open-source models

  for domain-specific adaptation
• Use

  transfer learning

  on small datasets for clinical use
• Build

  ensemble learning systems

  for diagnostic accuracy

🧬 Open-Source Model Adaptation

• Modify models from Hugging Face, Meta, or Google for

  medical understanding

• Build

  custom tokenizers

  for EMR/EHR and medical terminology
• Apply

  quantization, pruning

  , and other

  inference optimizations

• Develop custom

  loss functions

  ,

  training loops

  , and

  architectural variations

⚙️ MLOps & Deployment

• Create training and CI/CD pipelines using

  MLflow, Kubeflow, or W&B

• Scale distributed training across

  multi-GPU environments

  (Ray/Horovod)
• Deploy models with

  FastAPI/Flask

  , supporting both

  batch and real-time inference

• Monitor for

  model drift

  , performance degradation, and compliance alerts

🏥 Healthcare AI Specialization

• Build systems for:
• Medical text classification & entity recognition
• Radiology report generation
• Clinical risk prediction
• Auto-coding & billing
• Real-time care alerts
• Ensure

  HIPAA compliance

  in all stages of model lifecycle

✅ Required Qualifications💻 Technical Skills

• 4–6 years

  of Python development

• 2+ years

  in ML/AI with deep learning frameworks (PyTorch/TensorFlow)
• Experience modifying

  open-source transformer models

• Strong expertise in

  workflow orchestration tools

  (Airflow, Prefect, Dagster)
• Hands-on with MLOps tools (MLflow, W&B, SageMaker, DVC)

🔍 Core Competencies

• Strong foundation in

  transformers

  ,

  NLP

  , and

  CV

• Experience in distributed computing, GPU programming, and model compression
• Ability to explain and interpret model decisions (XAI, SHAP, LIME)
• Familiarity with

  containerized deployments (Docker, K8s)

🧰 Technical Stack

• Languages

  : Python 3.9+, CUDA

• ML Frameworks

  : PyTorch, TensorFlow, Hugging Face, ONNX

• Workflow Tools

  : Airflow, Prefect, Dagster

• MLOps

  : MLflow, Weights & Biases, SageMaker

• Infra

  : AWS, Kubernetes, GPU Clusters

• Data

  : Spark, Dask, Pandas

• Databases

  : PostgreSQL, MongoDB, Delta Lake, S3

• Versioning

  : Git, DVC

🌟 Preferred Qualifications

• Healthcare experience (EHR, medical NLP, radiology, DICOM, FHIR)
• Knowledge of

  federated learning

  ,

  differential privacy

  , and

  AutoML

• Experience with

  multi-modal

  ,

  multi-task

  , or

  edge model deployment

• Contributions to

  open-source projects

  , or

  research publications

• Knowledge of

  explainable AI

  and

  responsible ML practices

🎯 Key Projects You'll Work On

• Real-time

  clinical documentation automation

• Custom

  NER models

  for ICD/CPT tagging
• LLM adaptation for

  medical conversation understanding

• Real-time

  risk stratification pipelines

  for hospitals

🎁 What We Offer

• Comprehensive health, plans
• Flexible work options (remote/hybrid) with

  quarterly in-person meetups

📤 Application Requirements

1. Resume with ML/AI experience
2. GitHub or portfolio links (model code, notebooks, demos)
3. Cover letter describing your AI workflow or custom model build
4. Code samples (open-source or private repos)
5. Optional: Research papers, Kaggle profile, open challenges

🧪 Interview Process

1. Initial HR screening (30 mins)
2. Take-home Python + ML coding challenge
3. Technical ML/AI deep-dive (90 mins)
4. Model training/modification practical (2 hrs)
5. System design for ML pipeline (60 mins)
6. Presentation or walkthrough of past AI work (45 mins)
7. Culture fit + final discussion
8. References + offer

🏥 About the Role

AI that doesn’t just analyze data — it augments clinical decisions

Aarna Tech Consultants Pvt. Ltd. (Atcuality)

We believe in diversity, ethics, and inclusive AI systems for healthcare.