Machine Learning Engineer

ALKU; San Francisco (CA); On-site; USD 120,000 - 190,000; Full time

Job summary

ALKU is a Revenue Activation Platform. We are seeking a Machine Learning Engineer to own the end-to-end lifecycle of models powering our roleplay, scoring, and coaching products. You will train, fine-tune, and deploy models that operate on live sales-call data, including on-device deployments when latency and privacy demand it.

You will design and maintain evaluation frameworks and benchmarks, ensuring robust quality before releases and continuous improvement of model performance in production.

Qualifications

• Fine-tuning and production deployment of open-source models.
• On-device model deployment with quantization and distillation.
• Design and maintain evaluation frameworks and benchmarks.

Responsibilities

• Fine-tuning and production deployment of open-source models.
• On-device model deployment with quantization, distillation, and model compression.
• Evaluation infrastructure: benchmarks and regression suites to prove improvements before release.

Skills

Fine-tuning

Production deployment

On-device deployment

Evaluation infrastructure

Open-source models

Model lifecycle

Job description

The company is a Revenue Activation Platform: an agentic operating system for sales organizations. Rather than passively recording call activity, our platform actively shapes downstream outcomes - converting observed selling behavior into targeted roleplay simulations, individualized coaching, and workflow interventions that measurably improve rep performance without introducing additional management overhead.

In this role, you will hold end-to-end ownership of your models, from initial training through production deployment and ongoing operation. The majority of your work will center on fine-tuning and serving open-source models in production environments, deploying models on-device, and building the evaluation infrastructure that provides quantitative evidence of whether a change delivered a measurable improvement.

Position Summary

As a Machine Learning Engineer, you will design, build, and ship the models that power the company's roleplay, scoring, and coaching products: production systems operating on live sales-call data. This entails ownership of the complete model lifecycle: training and fine-tuning, production deployment, and sustained operational reliability post-launch.

Core Responsibilities

• Fine-tuning and production deployment of open-source models. You will identify and execute opportunities where self-hosted open-source models offer superior control over cost, inference latency, or model capability relative to third-party alternatives, and own their deployment and operation in production.
• On-device model deployment. Where customer latency or data-privacy requirements dictate, you will deploy models directly on-device, navigating the associated tradeoffs in quantization, distillation, and model compression while preserving output quality.
• Evaluation infrastructure. You will design and maintain the evaluation frameworks, benchmarks, and regression suites that provide statistically grounded evidence of model improvement or degradation, ensuring quality issues are detected internally before they reach customers.

Collaboration and Influence

You will work in close partnership with the founders and the broader engineering organization. Beyond execution, you will have substantive input into product and technical direction.