Specialist - AIML Software Development

Accelya; Mumbai; On-site; INR 2,500,000 - 5,000,000; Full time

Job summary

Accelya is seeking a senior AI/ML architect to own end-to-end AI/ML architectures for production products and platforms. You will lead evaluations, establish standards, and drive deployment strategies across classification, forecasting, and generative AI use cases.

You will guide deep learning and LLM initiatives, build scalable pipelines, and ensure model quality, governance, and reproducibility in a fast-moving aviation industry context.

Qualifications

• 3+ years of experience building and operating production AI/ML systems.
• Expert-level Python proficiency with strong software engineering fundamentals.
• Experience with ML frameworks such as Scikit-learn, PyTorch, TensorFlow, or Keras.
• Strong AWS experience across SageMaker, EC2, S3, Lambda, IAM, CloudWatch.
• Advanced Docker proficiency and containerized deployment experience.

Responsibilities

• Define and own end-to-end AI/ML architectures for business-critical products and platforms.
• Lead technical evaluations and design decisions for machine learning, deep learning, and generative AI initiatives.
• Establish engineering standards, model development practices, and deployment frameworks for AI systems.
• Drive architecture reviews and provide technical guidance on model selection, feature engineering, infrastructure, and deployment strategies.
• Architect and oversee production machine learning systems across classification, regression, forecasting, recommendation, optimization, and anomaly detection use cases.
• Lead advanced feature engineering, model selection, hyperparameter optimization, and model evaluation initiatives.
• Design robust experimentation frameworks for evaluating model performance and business impact.
• Establish standards for training, validation, testing, and deployment of machine learning models.
• Address challenges related to model drift, bias, explainability, fairness, and performance degradation in production.
• Drive adoption of best practices in statistical modeling and predictive analytics.
• Design and productionize deep learning solutions using CNNs, RNNs, LSTMs, Transformers, and modern neural architectures.
• Lead initiatives involving transfer learning, representation learning, self-supervised learning, and multimodal AI systems.
• Optimize model performance through pruning, quantization, distillation, and inference acceleration techniques.
• Evaluate and integrate state-of-the-art research advancements into production AI systems.
• Build scalable training and inference pipelines for large-scale deep learning workloads.
• Design and implement production-ready LLM-powered solutions including Retrieval-Augmented Generation (RAG).

Skills

Python

Scikit-learn

PyTorch

TensorFlow

Keras

Model deployment

Experimentation

MLOps

AWS SageMaker

CI/CD

Education

Master's Degree in Computer Science or related

Tools

Docker

Kubernetes

Terraform

AWS CDK

MLflow

Weights & Biases

Kubeflow

Job description

For more than 40 years, Accelya has been the industry’s partner for change, simplifying airline financial and commercial processes and empowering the air transport community to take better control of the future. Whether partnering with IATA on industry-wide initiatives or enabling digital transformation to simplify airline processes, Accelya drives the airline industry forward and proudly puts control back in the hands of airlines so they can move further, faster.

Roles & Responsibilities

AI/ML Strategy & Architecture Define and own end-to-end AI/ML architectures for business-critical products and platforms. Lead technical evaluations and design decisions for machine learning, deep learning, and generative AI initiatives. Establish engineering standards, model development practices, and deployment frameworks for AI systems. Drive architecture reviews and provide technical guidance on model selection, feature engineering, infrastructure, and deployment strategies. Produce technical design documents, architecture decision records (ADRs), and solution blueprints for AI initiatives. Evaluate emerging AI technologies and determine their applicability to business problems.

Machine Learning & Predictive Modeling

Architect and oversee production machine learning systems across classification, regression, forecasting, recommendation, optimization, and anomaly detection use cases. Lead advanced feature engineering, model selection, hyperparameter optimization, and model evaluation initiatives. Design robust experimentation frameworks for evaluating model performance and business impact. Establish standards for training, validation, testing, and deployment of machine learning models. Address challenges related to model drift, bias, explainability, fairness, and performance degradation in production. Drive adoption of best practices in statistical modeling and predictive analytics.

Deep Learning & Applied AI

Design and productionize deep learning solutions using CNNs, RNNs, LSTMs, Transformers, and modern neural architectures. Lead initiatives involving transfer learning, representation learning, self-supervised learning, and multimodal AI systems. Optimize model performance through pruning, quantization, distillation, and inference acceleration techniques. Evaluate and integrate state-of-the-art research advancements into production AI systems. Build scalable training and inference pipelines for large-scale deep learning workloads.

Generative AI & Large Language Models

Design and implement production-ready LLM-powered solutions including Retrieval-Augmented Generation (RAG) systems. Develop robust embedding, retrieval, ranking, and knowledge-grounding strategies. Lead model evaluation efforts focused on factuality, safety, latency, reliability, and cost optimization. Work with proprietary and open-source foundation models including OpenAI, Anthropic, Gemini, Llama, and Mistral ecosystems. Implement fine-tuning and adaptation techniques such as LoRA, QLoRA, and instruction tuning where appropriate. Collaborate with product teams to identify opportunities for integrating generative AI capabilities into business workflows.

MLOps, Infrastructure & Production Engineering

Define and drive MLOps strategy including model registries, experiment tracking, feature stores, and deployment pipelines. Design CI/CD processes for AI/ML workloads covering validation, testing, rollout, monitoring, and rollback mechanisms. Architect cloud-native AI infrastructure on AWS leveraging SageMaker, EC2, Lambda, EKS/ECS, S3, and CloudWatch. Ensure reproducibility and scalability of AI workloads through Docker, Infrastructure-as-Code, and modern deployment practices. Implement monitoring frameworks covering data quality, model drift, prediction performance, latency, availability, and system reliability. Establish operational SLAs and governance standards for production AI systems.

Required Skills

Core – Must Have 3+ years of experience building and operating production AI/ML systems. Expert-level Python proficiency with strong software engineering fundamentals. Deep understanding of machine learning algorithms, statistical modeling, model evaluation, and optimization techniques. Extensive experience with machine learning frameworks such as Scikit-learn, PyTorch, TensorFlow, or Keras. Strong expertise in deep learning architectures including CNNs, RNNs, LSTMs, Attention Mechanisms, and Transformers. Hands-on experience deploying machine learning solutions in production environments. Experience with Generative AI and Large Language Model applications including RAG architectures and embedding-based systems. Strong AWS experience across SageMaker, EC2, S3, Lambda, IAM, CloudWatch, and related services. Advanced Docker proficiency and containerized deployment experience. Practical experience with CI/CD pipelines and software delivery practices. Experience with MLOps platforms such as MLflow, Weights & Biases, Kubeflow, or equivalent. Strong understanding of experimentation frameworks, A/B testing methodologies, and model performance measurement. Excellent problem-solving, debugging, and technical decision-making skills. Good to Have Experience with distributed training technologies such as Ray, DeepSpeed, or PyTorch FSDP. Knowledge of vector databases including Pinecone, Weaviate, FAISS, ChromaDB, or pgvector. Experience with reinforcement learning, causal inference, or Bayesian modeling. Familiarity with Terraform, AWS CDK, Pulumi, or Infrastructure-as-Code tools. Exposure to multimodal AI systems involving text, image, and structured data. Experience with recommendation systems, ranking models, search relevance, or personalization engines. Contributions to open-source AI projects, technical publications, or research initiatives. Startup or product-company experience operating in fast-moving environments. Education Master's Degree (preferred) or Bachelor's Degree in one of the following disciplines, or equivalent professional experience: Computer Science Artificial Intelligence Machine Learning Data Science Statistics Mathematics Computer Engineering Related quantitative or computational fields Relevant certifications such as: AWS Certified Machine Learning Specialty AWS Solutions Architect Google Professional Machine Learning Engineer DeepLearning.AI Specializations are valued but are not substitutes for demonstrated engineering capability. Experience 4+ years of progressive hands-on experience in AI/ML Engineering, Machine Learning, Data Science, or Applied AI, with evidence of increasing technical ownership and leadership.

We are looking for evidence of:

End-to-end ownership of production AI/ML systems from problem definition through deployment and continuous improvement. Strong experience applying machine learning and deep learning techniques to solve real-world business problems. Experience building and maintaining production-grade AI platforms and services. Demonstrated technical leadership through architecture ownership, design reviews, mentoring, and engineering excellence initiatives. Experience collaborating with product, engineering, and business stakeholders to define and execute AI strategies. Ability to navigate technical ambiguity and convert complex business challenges into scalable AI solutions.

What does the future of the air transport industry look like to you?

Whether you’re an industry veteran or someone with experience from other industries, we want to make your ambitions a reality!

For more than 40 years, Accelya has been the industry’s partner for change, simplifying airline financial and commercial processes and empowering the air transport community to take better control of the future. Whether partnering with IATA on industry-wide initiatives or enabling digital transformation to simplify airline processes, Accelya drives the airline industry forward and proudly puts control back in the hands of airlines so they can move further, faster.