---
title: 'Member of Technical Staff, Post-Training, RL at Mirendil'
canonical: 'https://feeny.ai/job/member-of-technical-staff-post-training-rl-mirendil-san-francisco-ep9mzj7wn4a1'
type: 'job'
last_seen: '2026-09-11'
---

# Member of Technical Staff, Post-Training, RL at Mirendil

- **Company:** Mirendil
- **Location:** San Francisco, CA
- **Employment:** full-time
- **Work type:** onsite
- **Posted:** 2026-06-24
- **Last confirmed live:** 2026-09-11
- **Apply:** https://jobs.ashbyhq.com/mirendil/235e6bac-aad6-4caa-9427-1aff7190c45b

## Job description

Mirendil Mirendil is a tech-first company focused on solving core bottlenecks that unlock step-change acceleration across science and technology. Our first goal is to democratize frontier AI R&D across scientific disciplines. We are building a frontier AI research company and training our own models end-to-end.

## The Role

We are looking for research engineers to help build the post-training stack for frontier reasoning models. This role sits at the point where model capability, training dynamics, data, verification, and infrastructure all meet. You will design and run the experiments that turn a strong base model into a model that can solve difficult tasks reliably: choosing training objectives, shaping data mixtures, building verifiers, debugging reward signals, scaling runs, and understanding why a recipe works or fails. Researchers are also expected to have strong engineering skills. The best work here will involve both: forming hypotheses about training behavior, implementing them in real systems, running large-scale experiments, reading the resulting traces carefully, and turning the lessons into the next training run.

Some areas you may work on include:

- Post-training recipes: Develop and iterate on RL, SFT, and distillation recipes. Understand how choices in objectives, data mixtures, hyperparameters, rollout generation, and filtering affect efficiency, stability, capability, and final model behavior.
- Scaling RL: Make post-training work at larger scales: more tokens, longer trajectories, larger models, more steps, and larger compute budgets. This includes identifying the bottlenecks that appear only when an approach leaves the small-run regime.
- Long-horizon reasoning: Train models on tasks where success depends on many intermediate decisions. Develop methods for assigning useful feedback across long trajectories, where sparse rewards, credit assignment, exploration, and verification all become harder.
- Off-policy and asynchronous training: Work on training regimes where data is generated by older policies, different policies, or partially filtered policies. Build intuition and tooling for when off-policy data helps, when it hurts, and how to control the resulting instabilities.
- Verification and reward quality: Build robust verification pipelines for tasks where correctness can be checked automatically or semi-automatically. Detect and reduce reward hacking, false positives, brittle verifiers, and other failure modes that make RL look better than it really is.
- Multi-task post-training: Scale recipes across different task families and domains. Study the tradeoffs between specialization and generality, and design training mixtures that improve all capabilities together.
- Experiment analysis and debugging: Develop a deep empirical understanding of training runs. Diagnose regressions, separate real improvements from noise, design better ablations, and build the probes and analyses needed to make post-training less opaque.
- End-to-end execution: Work closely with systems, infrastructure, and data teams to get experiments from idea to production-scale runs. This includes making training pipelines reliable, ensuring data and verifier quality, and turning successful experiments into repeatable and scalable recipes.

If you're excited about building the infrastructure that makes frontier RL research possible at scale, we'd love to hear from you.

We offer a base salary of $300,000–$400,000 USD and a meaningful equity grant, depending on experience and background, along with competitive benefits.

## About Mirendil

## Company Overview
- **One-liner**: Mirendil is building an AI system that automates AI research and development, with the goal of creating a self-improving loop that accelerates scientific discovery.
- **Entity Type**: Private (Seed stage; raised $200M)
- **Headquarters**: San Francisco, California, USA
- **Founded**: Early 2026
- **Founders**: Behnam Neyshabur and Harsh Mehta

## Core Business
- **Primary industry**: Artificial Intelligence / AI Research & Development
- **Target customers**: B2B – initially AI researchers and engineers, but eventually scientists and domain experts (e.g., biology labs, drug discovery teams) who need frontier AI capabilities without building a full AI lab.
- **Mission**: “Democratizing frontier AI R&D to accelerate science and technology.” The company believes self-accelerating AI R&D is the most direct path to solving humanity’s pressing problems.

## Products & Services
- **Mirendil Platform**: A proprietary AI R&D system that trains frontier models specialized in AI research tasks (e.g., experimental design, hyperparameter search, model evaluation, code writing, debugging). The platform is designed to autonomously loop over research and engineering problems, controlling its own GPUs and improving over time with minimal human input. **Status**: Not yet shipped; under development.

## Market Standing
- **Valuation**: $1 billion (as of June 2026 seed round)
- **Key Metric (Funding)**: $200 million raised in seed round – one of the largest AI seed rounds ever.
- **Notable Investors/Partners**: Andreessen Horowitz (co-lead), Kleiner Perkins (co-lead), NVIDIA (participant).
- **Growth Signals**:
  - Raised $200M at $1B valuation just months after founding.
  - Founding team of 20+ researchers and engineers recruited from Anthropic, xAI, Google DeepMind, and OpenAI.
  - Founders Behnam Neyshabur (co-inventor of SAM optimizer, ex-Google, ex-Anthropic) and Harsh Mehta (ex-Google, ex-Anthropic, initiated automated AI R&D efforts) have deep frontier AI experience.
  - Targeted at compressing research cycles from months to days.

## Competitive Advantages
- **Self-Accelerating Loop**: The product is not just a model but a system that improves itself – “the loop is the product.” This creates a compounding advantage if it works.
- **Singular Focus**: The entire company is rebuilt from scratch around AI doing AI research, unlike traditional labs that use AI as a tool.
- **Top-tier Talent Density**: With only ~20 people, the team includes some of the most accomplished young researchers from the leading AI labs.
- **Strategic Investors**: Backing from a16z, Kleiner Perkins, and NVIDIA provides credibility, capital, and potential infrastructure access.

## Strategic Focus
- **Immediate goal**: Build the system that can autonomously perform AI R&D, making frontier research accessible to non-AI experts.
- **Long-term vision**: Democratize frontier AI R&D so that any lab (drug discovery, chemistry, biology, robotics) can leverage it without becoming a frontier AI lab themselves.
- **Current priorities**: Train models exceptional at AI research; build infrastructure for automated experimentation, evaluation, and iteration; keep the team small and high-compound.

## Why Work Here
- **Engineering culture**: “Small team, singular focus. We hire to compound, not to grow.” The team of ~20 operates with high autonomy and ownership.
- **Cutting-edge work**: Opportunity to work on one of the most ambitious AI projects – automating the very process of AI research.
- **Team composition**: Colleagues from Anthropic, xAI, DeepMind, OpenAI; founders known for foundational contributions (SAM optimizer, Blueshift/Mirerva).
- **Work environment**: Based in San Francisco; likely in-office given the intensity of the work (remote/hybrid policy not publicly detailed).
- **Notable perks**: Direct exposure to frontier model training and infrastructure; chance to shape a company from its earliest stage.

## Sources
1. [mirendil.com](https://mirendil.com/)
2. [mirendil.com/careers.html](https://mirendil.com/careers.html)
3. [kleinerperkins.com](https://www.kleinerperkins.com/perspectives/mirendil-building-the-system-that-builds-systems/)
4. [a16z.news](https://www.a16z.news/p/investing-in-mirendil)
5. [techfundingnews.com](https://techfundingnews.com/ex-anthropic-researchers-raise-200m-just-weeks-after-quitting-to-build-ai-that-creates-better-ai/)

## Other roles at Mirendil

- [Executive Assistant & Operations](https://feeny.ai/job/executive-assistant-operations-mirendil-san-francisco-c06e01p1zc49) — San Francisco, CA
- [Member of Technical Staff, Designer](https://feeny.ai/job/member-of-technical-staff-designer-mirendil-san-francisco-kzprmabdengf) — San Francisco, CA
- [Member of Technical Staff — Company Building](https://feeny.ai/job/member-of-technical-staff-company-building-mirendil-san-francisco-pgxgc7w41ejt) — San Francisco, CA
- [Member of Technical Staff, Inference](https://feeny.ai/job/member-of-technical-staff-inference-mirendil-san-francisco-sgh26y1tszjq) — San Francisco, CA
- [Member of Technical Staff, Model Evaluation](https://feeny.ai/job/member-of-technical-staff-model-evaluation-mirendil-san-francisco-xznefyyvx6s7) — San Francisco, CA
- [Member of Technical Staff, Post-Training, RL Environments](https://feeny.ai/job/member-of-technical-staff-post-training-rl-environments-mirendil-san-francisco-4xhaghhhy75v) — San Francisco, CA
- [Member of Technical Staff, Post-Training, RL Infra](https://feeny.ai/job/member-of-technical-staff-post-training-rl-infra-mirendil-san-francisco-szsxpdqdgqj8) — San Francisco, CA
- [Member of Technical Staff, Kernels](https://feeny.ai/job/member-of-technical-staff-kernels-mirendil-san-francisco-yjht6zkjbnqh) — San Francisco, CA
- [Member of Technical Staff, Pretraining](https://feeny.ai/job/member-of-technical-staff-pretraining-mirendil-san-francisco-7f3vnjtsrh49) — San Francisco, CA
- [Member of Technical Staff, Infrastructure](https://feeny.ai/job/member-of-technical-staff-infrastructure-mirendil-san-francisco-crx8w13g6s47) — San Francisco, CA