JOIN us
Applied AI Engineer
Location:
SF bay area, US.
New Resilience helps behavioral-health treatment centers lower client acquisition cost and increase admissions with an AI-native CRM, Alumni engagement tools, and a Referral Community that drives trusted, lower-cost referrals. We’re backed by respected operators in the space and used daily by admissions & outreach teams.
One step at a time, we’re on a mission to improve access to care using AI.
As an Applied AI Engineer, you will play a pivotal role in to building reliable, secure, agentic AI directly into our products. In this role, your focus will be on turning state-of-the-art techniques into tangible products. You’ll own the agent runtime and the systems around it: tool interfaces, permission-scoped data access, evaluation harnesses, observability, and safe execution. You’ll work closely with product and backend engineering to ship features end-to-end . This is a high-impact role where you’ll own projects from idea to production, and your contributions will directly shape our product and technical trajectory.
You will collaborate with a world-class team to build innovative, scalable systems. As part of our fast-paced startup, you’ll have the opportunity to influence technical direction, contribute ideas across the stack, and rapidly grow into taking more leadership as we scale. We value first-principles thinking and a get-things-done attitude – you will have broad autonomy to experiment and iterate in pursuit of the best solutions.
What you’ll do
Drive AI Innovation: Research, prototype, and build cutting-edge product, and customer facing AI Systems, pushing the state-of-the-art to solve our core technical challenges.
Build agentic workflows in production: Design and implement AI agents that can take actions across workflows (e.g., follow-ups, task creation, data enrichment, summaries, etc.), with clear boundaries and human-in-the-loop controls where needed.
Own the tool and action layer: Define strict tool/function contracts (schemas, validations, idempotency) so the model can safely query and update CRM data without hallucinating fields or performing unsafe actions.
Security and privacy by design : Enforce least-privilege access, row-level permissions, redaction, audit logging, and safe handling of PHI across prompts, logs, and integrations.
Reliability, evaluation, and monitoring: Build evaluation suites for agent success (task completion, correctness, groundedness, safety), set up regression tests, and instrument production monitoring for failures, latency, and quality.
Iterate fast with feedback loops: Turn user feedback and real usage data into better prompts, workflows, retrieval, and guardrails. Improve quality without sacrificing safety.
Contribute to core backend architecture: Collaborate on APIs, async jobs, background processing, and data model changes needed to support agentic features at scale.
You may be a fit if you have
Applied AI experience: You’ve shipped LLM-powered features into production and you know how to make them reliable: tool use, retries, fallbacks, monitoring, and safe rollouts. Solid grounding in modern deep learning and LLM systems, including test-time compute/inference-time scaling, training and eval best practices, and data/feedback loops (e.g., self-supervised learning, synthetic data, active learning, automated data collection).
Strong backend engineering: 2+ years of backend engineering experience shipping production services (APIs, data models, background jobs). Strong Python skills and comfort in production services (APIs, async workers, background jobs). GCP experience building and running services on Google Cloud (e.g., Cloud Run or GKE, Cloud SQL, Pub/Sub/Cloud Tasks, Logging/Monitoring).
Agent/tooling fundamentals: Experience with function/tool calling, schema enforcement (JSON schema/Pydantic), workflow orchestration/state machines, and preventing destructive actions (approval gates, dry-runs, allowlists).
Security mindset (privacy / compliance friendly): You’ve worked on systems with sensitive data and understand permissioning, audit logs, redaction, secrets management, and how to avoid leaking data through prompts/logs.
Evaluation discipline: You know how to define “correct,” build test sets, run regression evals, and measure improvement over time.
Clear communicator, low-ego collaborator: You explain tradeoffs clearly, move quickly, and enjoy building in a fast-paced environment.
Why join us
Build category-defining agentic workflows: for a trust-based industry that needs better tools
High ownership: you’ll shape the agent architecture and what “autonomous” means in a real product
Mission: help treatment centers help more people, sustainably.
Benefits
Health, Dental, and Vision Insurance.
15 Paid Holidays.
Gym membership covered.
Frequent team dinners, events, and off-sites.
Email aaraddhya@newresilience.ai with your resume/LinkedIn and a short note on a complex problem you tackled and solved.
