Build labeled evaluations from real agent activity and publish monitor precision and recall.
Run monitors across customer fleets and identify unread events.
Turn patterns into proposed policies and refinements for security engineers.
Sondera develops a policy engine for AI agents that enforces natural-language rules to ensure agents act as intended. The team includes veterans from top security and AI companies and is backed by leading venture firms, emphasizing a remote, writing-intensive culture.
You own the architecture of Sondera's backend services, deciding how they are designed, scale, and fail closed.
You set the technical bar, run design review, and turn AI security research into enforceable systems-level guarantees.
In your first year, you write the architecture of record, prove verdicts replay exactly, and raise the bar on every principal hire.
Sondera AI agents act on what people write, enforcing policy through natural-language rules to ensure semantic integrity. Backed by Decibel, Runtime Ventures, and Aviso Ventures, their team has built products at Cylance, Databricks, and HiddenLayer.
Deploy Sondera through customers' device management and integrate their agents, turning written rules into policy.
Own each customer's post-deployment success, including upgrades, policy tuning, and adversarial testing.
Collaborate with GTM to drive technical evaluations and expansions, bringing recurring needs back into the product.
Sondera builds AI agents that enforce written policies on agent actions, using natural-language rules. Their experienced team is backed by Decibel, Runtime Ventures, and Aviso Ventures, fostering a remote-first culture.
Own the services that render verdicts on AI agents' actions and maintain the record of agent trajectories for enterprise security teams.
Keep trajectory reads fast at scale, design event models for all agent surfaces, and ship schema changes without production downtime.
Ensure every verdict cites exact events behind it for customer audit records.
Sondera works on the open problem of semantic integrity for AI agents, ensuring every action carries out what an authorized person meant by autoformalizing plain language rules into policy. Our small team has built products at Cylance, Databricks, and other security companies, backed by notable venture firms.
Own the API service behind the Sondera console that serves all clients.
Design and implement versioned APIs, authorization models, and staged rollouts.
Turn AI security research into enforceable systems-level controls.
Sondera builds AI agents that enforce policies based on written rules to prevent harmful sequences. The team has experience at companies like Cylance and Databricks, and is backed by Decibel, Runtime Ventures, and Aviso Ventures.
Design and build the agent runtime, including dispatch, execution lifecycle, tool calling, and error handling.
Build the extension system and trust boundary, ensuring safe default actions and auditable permissions.
Design the context system and inference layer, with provider abstraction and observability.
Yaklabs is building a desktop AI application that turns knowledge workers into AI power users. The small, high-intensity team is funded and led by industry veterans.
Own quality for non-deterministic, multi-agent AI systems, building the evaluation harness that underpins AI governance.
Verify grounding and faithfulness, building automated checks to stop hallucinations from reaching production.
Red-team for prompt injection, jailbreaks, and tool misuse, and prove multi-tenant data isolation.
SOFICO Intelligence uses AI to extract value from back-office systems, building agentic AI for financial, leasing, fleet, and mobility data. Backed by global ERP leader SOFICO, it's a remote-first team of curious, self-directed builders.
Own the context layer: manage what Mimo knows, when, and at what cost, with retrieval and memory as one budget.
Define tool contracts and build evaluation practice: schemas, results, failure modes, and regression gates for every change.
Shape how Mimo reasons, asks, explains, and recovers, and decide which model runs where.
MimoLabs builds Mimo, an agentic system for making physical things, from design to production across 3D printing, CNC, and laser cutting. We are a small, VC-funded, experienced team based in Detroit and remote across the US, fast, adaptable, and honest.
Plan, build, and improve background-agent infrastructure that enables reliable, trackable, and repeatable automated engineering work.
Develop systems for evaluating agent performance using task outcomes, quality measures, human intervention, cost, and other signals.
Improve how engineers discover, understand, and use shared AI tools and workflows, and create clear guidance and documentation.
Vanta helps businesses earn and prove trust through automated security monitoring and a leading Trust Management Platform. Founded in 2018, the company has a kind, inclusive, and talented team, and thousands of companies rely on Vanta for real-time security and compliance.
Partner with Platform Engineering, Infrastructure, and Developer Productivity to integrate agent capabilities.
Reddit is a community of communities built on shared interests, passion, and trust, with 100,000+ active communities and ~130 million daily active users. We are one of the internet's largest sources of information, fostering open and authentic conversations.
Design evaluations of conceptual reasoning in AI safety, covering governance, alignment, macrostrategy, and economics.
Develop rubrics that separate genuine reasoning from plausible-sounding filler.
Iterate on tasks with frontier models and collaborate with a top research team.
Trajectory Labs is a startup automating AI safety research. We build safety and alignment evals, red-teaming programs, and RL environments for frontier AI labs. As a one-year-old company, we've been featured in major frontier lab safety reports and led high-impact red-teaming efforts.
Architect and build production multi-agent systems for safe operation across customer business tools.
Own the MCP connecting Distru to AI assistants, ensuring reliability and observability.
Build eval suites and a contextual memory system to improve agent performance from real customer use cases.
Distru is the #1 ERP in the cannabis industry, having processed $15B in sales through its platform. They are a growing, profitable company with a 5/5 Glassdoor rating, backed by investors like Felicis Ventures and Global Founders Capital, and operate with a culture focused on AI innovation and mutual respect.
Build evaluation frameworks for AI agents, including datasets and metrics for real-world task performance.
Design experiments to compare models and tools, identifying regressions and trade-offs.
Investigate agent failures and improve performance across models, harnesses, and tooling.
Gauntlet builds the financial systems of the future, operating across the onchain finance stack to offer vault products. They serve over $1.7B in client TVL across fintechs, protocols, exchanges, and capital allocators, combining traditional finance and crypto-native expertise.
Build post-training and evaluation systems for multimodal agents operating browsers and desktops.
Turn real workflows into environments, rewards, and experiments that drive reliable agent improvement.
Work directly with customers and research partners while maintaining a substantive research agenda.
Prime Intellect is building the open superintelligence stack, providing infrastructure for post-training frontier AI models. The company has raised $150M in funding and fosters a culture of ownership and collaboration for ambitious AI teams.