5 months ago
Bengaluru, IndiaSenior
Responsibilities
- Build and maintain high-performance distributed systems for agreement processing, understanding, and creation.
- Design multi-agent frameworks with state management, reliability, and long-running autonomous workflows.
- Architect Retrieval-Augmented Generation pipelines and context-management strategies.
- Build evaluation frameworks, golden datasets, domain-specific evaluation sets, and offline/online experiments.
- Author, version, and optimize production prompts with defenses against prompt injection.
- Execute end-to-end model fine-tuning for tailored agreement use cases.
- Implement ML pipelines and CI/CD workflows with observability, telemetry, log parsing, and deployment of generative AI services.
- Collaborate with Applied Science and Product Management to deliver AI capabilities into production features.
Requirements
- 5+ years of experience in machine learning engineering, software engineering, or related operational roles.
- Proficiency in Python design patterns, asynchronous programming, and performance optimization.
- Experience deploying and managing containerized ML services using Kubernetes.
- Experience building high-performance, scalable distributed systems for large-scale processing and real-time user experiences.
- Understanding of data ingestion, model training, and production-grade monitoring across the ML lifecycle.
- Direct experience building with large language models and implementing complex prompt engineering.
- Experience deploying and maintaining ML models in high-traffic production environments.
- Preferred: cloud-native development experience with Azure services.
- Preferred: experience designing agent-loop architectures involving tool use, self-correction, and multi-step reasoning.
- Preferred: ability to own evaluation systems and create golden and domain-specific evaluation datasets.
- Preferred: experience with distributed task queues or stateful workflow engines for complex AI processes.
Benefits
- Hybrid role requiring a minimum of two days per week in the office, with frequency subject to team needs.
- Docusign provides reasonable accommodations for qualified individuals with disabilities and religious accommodations during the application process.
- Employees may opt out of AI-assisted interview recording or transcription without impact on candidacy.
Tech Stack
Categories
About DocuSign
DocuSign builds e-signature, contract lifecycle management, and agreement workflow software used by businesses to prepare, sign, and manage contracts. It sells cloud-based subscriptions and APIs that integrate with systems like Salesforce, Microsoft, and Google to automate document workflows and compliance. Founded in 2003 and headquartered in San Francisco, DocuSign is a public company on NASDAQ with over 1.5 million customers in more than 180 countries.