Software Engineer IV

What You Will Do

Forward Deployed Engineering

• Embed within product teams to deliver end-to-end AI/ML integrations - API orchestration, workflow management, and error handling - directly inside the host team's codebase and release processes.
• Produce structured handover artefacts for every engagement: merged PRs, architecture notes, runbooks, observability dashboards, and knowledge transfer sessions.

Integration & Pipeline Engineering

• Implement and extend Cognition's integration patterns - coordinator/extractor/mailbox services, Kafka pipeline patterns, and RESTful API contracts.
• Own and scale the Cognition v3 architecture - data pre-processing, post-processing, the audit sub-system, and the Conduct integration pipeline - handling high message volumes with reliable throughput.
• Instrument services with Prometheus and Grafana; maintain observability across Logic's owned components.

Key Initiatives (2026)

• OCRaaS at scale - deliver the OCR-as-a-Service integration for eDiscovery and Conduct.
• Mature as-a-Service patterns to simplify access to Cognition AI/ML capabilities across the product estate.
• Enhanced attachment processing - extend the Cognition ingestion pipeline to handle larger attachments at higher throughput, removing current size and volume constraints for customers.
• Cognition v3 reliability - harden the pre/post-processing communication surveillance pipeline for production resilience, operational visibility, and zero-surprise deployments.

AI Enablement & Engineering Standards

• Champion and actively use AI productivity tooling (Windsurf, Claude Code) and embed these practices in the teams you work with.
• Write clean, well-tested Kotlin (or Java); contribute to architecture and design reviews across Logic and host teams.

What We Offer

Smarsh exists to protect the truth - preserving the integrity of communications across the world's most regulated industries. The AI you build here does not power recommendation feeds or ad targeting. It runs inside the compliance workflows of Tier 1 banks, global asset managers, and regulated enterprises where auditability, accuracy, and trust are non-negotiable.

You will work at the frontier of that mission - integrating AI and moving into agentic AI to solve problems that have no off-the-shelf answer. The constraints of highly regulated environments make the engineering harder and the outcomes more durable. That combination is rare.

• Scope - your work spans Conduct, eDiscovery, and Enterprise Archive, with direct impact on the capabilities Smarsh brings to its most strategic customers.
• Visibility - sustained interaction with senior engineering and product leadership, and a seat at decisions that shape platform direction.
• Progression - a clear path toward Lead Engineer, Architect, or Engineering Manager within Cognition Engineering, in an organisation that treats engineering craft seriously.