Founding Member of Technical Staff (MTS)

VizopsAI; San Francisco (CA); On-site; USD 150,000 - 220,000; Full time

Job summary

VizopsAI, based in the San Francisco Bay Area, is searching for a Founding Member of Technical Staff (MTS) to create secure production-grade systems that enhance AI agents. You will collaborate with customers to turn their objectives into actionable signals, emphasizing system reliability and performance.

The ideal candidate should possess strong Python programming capabilities, experience with backend systems, and a knack for working with cloud services like GCP and AWS. This full-time role offers competitive compensation, including equity.

Qualifications

• 2+ years experience building ML or backend systems.
• Proficient in building production-grade systems.
• Familiarity with cloud services (GCP/AWS).

Responsibilities

• Build backend services for training, evals, telemetry, and online policy updates.
• Instrument observability for optimization loops.
• Collaborate with product & customers for measurable improvements.

Skills

Strong programming in Python

Containerization (Docker/K8s)

Fluency with data & infra

Debugging production systems

Tools

GCP/AWS

Job description

# Founding Member of Technical Staff (MTS)Bay Area, CAFull-time$150k-$220k + equity## About UsVizopsAI is the secure runtime for custom enterprise software. We provide the production layer that turns AI-generated internal tools into compliant, hardened applications — wrapping raw AI code in enterprise-grade identity, security, and infrastructure best practices. We're a lean, fast-moving team building the industrialization layer for the AI app revolution.We're an early stage venture-backed AI-native startup based in the SF Bay Area. The founding team combines deep AI/ML research leadership at Google DeepMind, Amazon Alexa, and Oracle Cloud with AI Product leadership at Verkada, AWS and Sony. Technical leadership includes PhDs from Johns Hopkins specializing in deep learning and optimization.We already have multiple customers locked in and are bringing on rockstars to build the infrastructure that makes enterprise AI adoption safe and scalable.## About the RoleAs a Member of Technical Staff (MTS), you'll build production-grade systems that power continuous optimization loops for AI agents—from evaluation pipelines and data/trace infrastructure to APIs that deploy improved policies. This role is a blend of MLE + backend engineering with a strong customer empathy component. You'll partner closely with customers and products to translate real-world objectives (accuracy, latency, cost, safety) into measurable signals and reliable services.**Note:** Unlike our Founding AI Engineer role, there's no expectation to read/implement research papers from scratch—the bar for engineering rigor, ownership, and ambiguity-handling remains high. You'll need to demonstrate clear communication and high ownership in a fast, evolving environment.## What You'll Do* •Build backend services for training, evals, telemetry, and online policy updates* •Instrument observability to make optimization loops inspectable and reliable* •Translate customer KPIs into reward signals, guardrails, and success metrics* •Collaborate with product & customers to reduce time-to-uplift and land measurable improvements in production* •Scale distributed workloads for training/serving. Improve reliability, cost, and latency over time## What We're Looking For* •Strong programming in Python; comfortable with backend systems* •You've shipped production systems and can debug other people's code* •Fluency with data & infra - Containerization (Docker/K8s), cloud (GCP/AWS)* •2+ years experience building ML or backend systems## Nice to HaveSome of the libraries are very new (as of Nov 2025) and we don't expect people to know them already* •Experience with RL, reward modeling, LLM evals, or agent stacks (retrievers, tool routers, orchestration)* •Familiarity with vLLM, LangSmith/Langfuse, SkyRL, Verl, Llama Factory, Agent Lightning* •LLM post-training exposure (preference data collection, safety/guardrails, structured evals)