Full Stack AI Engineer

Multiscale AI
Atlanta, GA · Full-time

The Role

Multiscale AI builds AI agents and digital twins that help semiconductor fabs optimize their manufacturing processes. We are hiring a hands-on senior engineer to design, build, and operate the agent infrastructure behind MIND, our agentic platform: the systems that run long analyses against customer-hosted data, keep them reliable under concurrent use, and make their results trustworthy enough for a process engineer to act on.

This is an individual-contributor role. You will write production code in our codebase from your first week and own your subsystems end to end.

Responsibilities

• Design and build durable orchestration for long-running agent jobs: chunked, checkpointed, idempotent execution that resumes after failure and handles mid-run changes to the request
• Build the data-access layer between agents and customer-hosted databases: parameterized queries pushed down to the store, outbound-only connectors, pre-aggregation, and per-tenant quotas that protect shared production systems
• Implement isolated, resource-limited execution environments for agent-generated analysis code
• Implement credential brokering for agents: short-lived, scoped credentials with user identity passed through, so customer access policies apply and secrets never enter model context
• Build multi-agent workflows where appropriate, and deterministic services where they are not
• Build human-in-the-loop controls: cost-threshold confirmations, review tooling, and judge-gated release flows
• Instrument the system end to end: traces tied to job IDs, provenance stamped on every result, and evaluation suites run as release gates
• Develop MCP servers, tool-registration APIs, and the API contracts between the agent layer and the core platform
• Operate what you build: on-call participation, incident response, and post-incident fixes

Required Qualifications

• 5+ years of production software engineering, including 2+ years building LLM or agentic systems
• Hands-on production code within the last 18 months, in an existing codebase, including migrations or refactors under live traffic
• Experience designing a distributed system from the ground up: queues, workers, timeouts, retries, and failure handling, with the ability to explain the trade-offs
• Experience operating systems you built, including on-call responsibility and incident resolution
• Strong Python; proficiency with Docker and Kubernetes on a major cloud (Azure, AWS, or GCP)
• Working knowledge of LLM evaluation: golden datasets, LLM-as-judge, regression gating
• Understanding of LLM fundamentals (context handling, inference cost, failure modes) sufficient to reason about latency, memory, and reliability
• Judgment about when an agent is the right tool and when deterministic code is

Preferred Qualifications

• Durable-workflow engines such as Temporal, or an equivalent built in-house
• Query pushdown, read-only connectors, or quota management against customer-owned or regulated data stores
• Secrets management and identity passthrough for automated systems
• Graph databases (Neo4j or similar) in production
• Semiconductor manufacturing, fab operations, tool sensor data, or materials science

Interview Process

• Intro call (15 min) with a short design question
• Technical coding round (60 min): hands on coding
• Technical system design round (60 min): fundamentals and a from-scratch systems design
• Decision within days of the technical round

Travel Requirements

Travel to customer sites (roughly 15–20%)

How to Apply

Interested candidates should submit their resume along with a brief note describing a production system they personally built or materially changed in the last 18 months, highlighting one scalability challenge and how they addressed it, to jobs@multiscale.ai with the position title in the email subject line.

Lead DevOps / Platform Engineer

Multiscale AI
Remote (US-friendly) | Full-time

The Role

We are hiring a Lead DevOps / Platform Engineer to own cloud-native infrastructure, platform engineering, and the operational backbone of Multiscale’s AI systems. This person will drive Kubernetes platform maturity, GitOps, internal developer platforms, multi-cloud reliability, and production AI agent systems (MCP, RAG, multi-agent workflows).

The role sits at the intersection of platform engineering and AI infrastructure: building the systems that let our teams ship reliably while also designing and operating the agentic tooling that automates operations themselves. Ideal candidates have deep hands-on experience shipping production Go services on Kubernetes, multi-cloud platforms, and production AI agents.

Responsibilities

• Lead cloud-native modernization and platform engineering standards across teams; set coding, documentation, and operational standards.
• Own Terraform / Infrastructure-as-Code initiatives that eliminate configuration drift and dramatically reduce provisioning time.
• Architect and execute migrations (e.g., Docker Compose → Kubernetes) with progressive delivery, automated rollback, and strong security posture.
• Design and operate private multi-tenant Kubernetes clusters with RBAC, hardening, and secrets management (HashiCorp Vault or equivalent).
• Build and evolve an Internal Developer Platform with self-service infrastructure and GitOps (Argo CD or equivalent) for declarative delivery and automated drift correction.
• Establish and run organization-wide CI/CD with approval workflows, team-based deployments, static analysis, and automated testing gates.
• Design monitoring, alerting, and SLAs (Prometheus, Grafana, Datadog or equivalent) with anomaly detection and proactive incident response.
• Lead development of AI agents on the Model Context Protocol (MCP) for operational automation, self-healing remediation, and workflow orchestration.
• Mentor engineers, drive engineering excellence, and partner closely with product and research teams on reliability and platform needs.

Required Qualifications

• 10+ years of software engineering experience with a strong focus on distributed systems, cloud-native platforms, and DevOps/platform engineering.
• Deep production experience with Go (preferred) or equivalent systems languages building microservices on Kubernetes across AWS, Azure, and/or GCP.
• Hands-on ownership of Kubernetes platforms: multi-tenant clusters, progressive delivery, service mesh (Istio or equivalent), Helm, and security hardening. • Proven Terraform / IaC work that has measurably reduced provisioning time and eliminated drift.
• Experience building or operating production AI agent systems, RAG pipelines, vector search, and/or Model Context Protocol (MCP)-based tooling.
• Strong CI/CD design and operation (GitHub Actions, Jenkins, Argo CD, etc.) with security and quality gates.
• Experience with observability stacks (Prometheus, Grafana, Datadog, ELK) and defining/operating SLAs.
• Track record of mentoring engineers and driving platform standards in a fast-moving environment.
• Bachelor’s degree in Computer Science or equivalent practical experience.

Preferred Qualifications

• Co-founding or early-engineer experience at a cloud-security, CNAPP/DSPM, or platform startup.
• Experience building multi-agent systems with structured state, human-in-the-loop checkpoints, and production RAG/vector search.
• Prior work migrating large monoliths to microservices or building Internal Developer Platforms.
• Familiarity with semiconductor manufacturing, physics-informed AI, or digital-twin systems (nice-to-have; not required).
• Experience with HashiCorp Vault, multi-cloud cost optimization, and compliance-oriented security scanning (CIS, SOC 2, etc.).

What Success Looks Like

• Infrastructure provisioning time reduced by 50–80% with near-zero configuration drift.
• Reliable, self-service platform that development teams trust for production deployments.
• Production AI agents handling meaningful operational workflows with clear safety and approval boundaries.
• Measurable improvements in deployment frequency, MTTR, and platform security posture.
• Clear engineering standards and mentorship that raise the quality bar across teams

Multiscale AI builds physics-informed AI and digital-twin systems for semiconductor manufacturing. We value precise, high-signal engineering, clear ownership, and systems that are both powerful and reliable.

How to Apply

To apply, interested candidates may send their resume to jobs@multiscale.ai with the position title in the subject line.