Cantina Labs

• Produce and direct scaled creative across all performance marketing surfaces (Meta, TikTok, YouTube, mobile ad networks, etc) with multi-million dollar ad spend. • Develop performance creative concepts built on intimate understanding of Cantina as a product/platform and its consumer use-cases. • Build and own processes for producing high volume, iterative ads leveraging both traditional and AI-generated production. • Define creative strategies and vision for AI-generated ads. • Own creative performance analytics to track what's working and iterate on best-performing concepts and campaigns. • Set the creative inputs across creative testing plans to optimize creative. • Manage external agencies and cross-functional teams to ensure high-quality creative output across all performance marketing channels. • Lead regular creative brainstorms, reviews, and presentations to internal stakeholders, including leadership. • Oversee the QC process for in-house and external creatives: Create scalable creative review processes and quality standards that maintain excellence at high volume. • Stay up-to-date on and apply emerging technologies, production methods, and best practices in performance creative and performance marketing.

• Designing model evaluation pipelines for models in development and production • Designing user studies for subjective model evaluations. • Converting requirements into measurable metrics. • Designing and developing automated evaluation dashboard to see model performances and compare results. • Training new models to capture new and different evaluation metrics. • Communicating with the model team to help design better models based on the evaluation results. • Communicating with the data team to help decide the type of data necessary to improve model performance. • Communication with the product-manager to make sure product requirements are correctly measured. • Help grow the evaluation team as the founding member. • Lead the evaluation team in the future.

• Dataset ownership: define specs; audit and curate large-scale audio/text; close corpus gaps and fix sample-level issues. • Quality instrumentation: build automated gates/metrics (e.g., SNR, clipping, VAD, WER, SV/LID, safety) with dashboards; validate against listening tests. • Classifiers and filters: train lightweight models to tag, score, and filter data (VAD, ASR gating, LID, SV/diarization, noise/safety); calibrate to subjective outcomes. • Cleaning and integrity: apply denoise/dereverb/de-clip when beneficial; deduplicate and decontaminate; prevent leakage; maintain lineage and versioned releases. • Data selection: optimize mixtures via sampling, weighting, curriculum, and active learning; mine hard negatives and long-tail cases. • Tooling and pipelines: ship reproducible ETL and validation; integrate quality gates into training/eval; add monitoring and alerts. • Human-in-the-loop and compliance: run MTurk/vendor annotation with strong QC; ensure consent/licensing/policy compliance; collaborate across teams and document datasets.

• Designing model evaluation pipelines for models in development and production • Designing user studies for subjective model evaluations. • Converting requirements into measurable metrics. • Designing and developing automated evaluation dashboard to see model performances and compare results. • Training new models to capture new and different evaluation metrics. • Communicating with the model team to help design better models based on the evaluation results. • Communicating with the data team to help decide the type of data necessary to improve model performance. • Communication with the product-manager to make sure product requirements are correctly measured. • Help grow the evaluation team as the founding member. • Lead the evaluation team in the future.

• Build and maintain data pipelines for large video generation models, including data ingestion, parsing, filtering, preprocessing, and dataset curation at scale, using tools such as AWS S3 and DynamoDB. • Design and run annotation workflows across platforms such as MTurk, Prolific, and Mechanical Turk, including task design, quality control, and label validation. • Train, evaluate, and improve smaller supporting models used for data filtering, quality assessment, preprocessing, or other parts of the ML pipeline. • Partner closely with research and engineering teams to turn experimental workflows into scalable, repeatable systems that support model training and evaluation. • Own data quality across the pipeline by identifying bottlenecks, failure modes, and low-quality sources, and continuously improving tooling and processes. • Build internal tools and automation that make it easier to prepare datasets, launch annotation jobs, monitor outputs, and support model development end to end. • Drive larger pipeline projects from start to finish, such as new dataset creation efforts or upgrades to labeling and preprocessing infrastructure. • Work within a Kubernetes-based training infrastructure, ensuring datasets are properly prepared, formatted, and delivered to training clusters. • Profile and optimize research model inference scripts used in preprocessing steps, ensuring that model-driven filtering and transformation stages run within practical time and cost constraints when applied to large-scale raw data.

About Cantina Cantina is building an agentic security operating system that spans application security, security operations, and agent security. We believe the next generation of security products should do more than aggregate alerts or automate isolated tasks. They should understand context, reason across systems, help teams investigate what matters, and safely take action. This is still an emerging space. Many of the most important risks, design constraints, and product opportunities haven’t been discovered yet. We need people who can help us build the product while also uncovering the unknown unknowns that come with combining security systems and agentic AI. The Role We’re hiring a security engineer who wants to build products in the AI era. You’ve spent years understanding how security teams actually work—how incidents get triaged, how alerts get tuned, how detection logic gets written and maintained, how appsec findings get prioritized. Now you want to build the product you wish existed. This is not a security review role, and it’s not a generic backend engineering position. We need someone whose core instincts come from security—understanding attacker behavior, operational failure modes, what actually matters when a SOC is under pressure—and who can turn that knowledge into product. The AI and product engineering dimensions are real parts of the job, but they’re the growth opportunity, not the entry requirement. If you have strong systems engineering skills and genuine curiosity about how agents, tools, and orchestration work, you’ll learn the rest here. What You’ll Do - Build product capabilities across application security, security operations, and agent security - Turn real security workflows into product experiences and platform primitives - Design systems that ingest, correlate, triage, and act on security signals - Help define safe patterns for agents, tools, permissions, memory, and execution boundaries - Identify hidden risks and failure modes that only someone with real security experience would see - Partner with product and engineering to make strong tradeoffs between speed, usability, and security - Contribute to evaluation, testing, observability, and guardrails for agentic behavior - Raise the team’s overall understanding of security architecture, operations, and AI risk What You Bring The non-negotiable: - Deep experience in one or more of: security engineering, application security, detection engineering, incident response, security operations, or security platform engineering - Strong hands-on experience building and shipping software—you write code, not just review it - The ability to reason clearly in ambiguous spaces and surface risks early Highly valued but learnable here: - Experience with AI/LLM application architecture, agent frameworks, or orchestration systems - Product judgment—translating messy technical workflows into usable product decisions - Comfort working across technical and non-technical teams Relevant Background You’ve likely worked with systems and workflows like these: - SIEMs: Splunk, Elastic, Microsoft Sentinel, Chronicle, Panther, or similar - EDR/XDR: CrowdStrike Falcon, SentinelOne, Microsoft Defender, or similar - SOAR / Automation: Tines, Torq, Cortex XSOAR, or similar - Appsec tooling: Semgrep, Snyk, CodeQL, Burp Suite, Wiz, or similar We don’t expect experience with every tool above. We want someone who has been close enough to these environments to understand how modern security teams investigate, prioritize, and respond. Technical Environment - TypeScript / Node.js (primary stack—willingness to work in this is required, prior experience is preferred) - API and integration-heavy systems - Backend and distributed systems design - Security data models, workflow design, and systems integration Why This Role Is Different Most security product companies hire engineers and teach them security, or hire security people and limit them to advisory roles. We’re looking for someone who can do both: ship real systems and bring the security depth to see what others will miss. You’ll have real influence over what gets built and how. If you’ve been frustrated by security products that clearly weren’t built by anyone who’s actually worked in security, this is your chance to fix that.

• Build and maintain end-to-end data pipelines for large-scale image and video datasets: collection, filtering, augmentation, conditioning alignment, and efficient storage/sampling. • Implement model architectures (diffusion, autoregressive, flow-based, diffusion transformers, etc.) and maintain high-throughput PyTorch training loops for large-scale image and video diffusion models. • Run and manage large-scale training experiments on multi-GPU and multi-node setups (DDP, FSDP, DeepSpeed). Debug training instabilities, loss spikes, and convergence issues. • Apply quantization, pruning, and knowledge distillation techniques to compress models without sacrificing quality. • Collaborate with researchers and translate state-of-the-art research papers into working implementations in our internal codebase (e.g., new attention mechanisms, sampling schedules, or conditioning methods). • Build and maintain evaluation pipelines of image quality, video consistency, and perceptual metrics. • Set up and maintain human annotation and evaluation pipelines using services like AWS GroundTruth. • Profile and optimize training speed, GPU memory utilization, and iteration time. Implement inference optimizations to reduce latency and compute cost. • Work with acceleration toolchains such as torch.compile, Triton, TensorRT, or ONNX where appropriate

• Evaluate new image generation and identity preservation papers and models. • Develop and deploy new versions of the image generation and image analysis pipelines • Monitor and fix production issues that impact users • Fine-tune and optimize models to improve character consistency, prompt responsiveness, and inference latency • Design and run experiments to benchmark model performance, tracking quality metrics across generations of pipeline improvements • Collaborate with cross-functional teams to translate product requirements into ML solutions and bring new generative features from prototype to production