Founding AI/ML Engineer

first dedicated AI/ML hire

that make our platform selfoptimizing.

What you’ll build & own

Horizon Highimpact work

03 months

• Stand up modelserving infra (Vertex AI or GKE) for policy nets.

• Productionize CUPED variancereduction pipeline in BigQuery ML.

• Pair with founder to feed reward signals into the agentic runtime.

39 months

ReinforcementLearning Budget Optimizer

Sampling → PPO).

• Automate syntheticcontrol jobs on Vertex AI for geolocked campaigns.

• Build a feature store merging offline postal scans & streaming web

events.

918 months

Finetune LLMs with LoRA/QLoRA

generation and craft robust prompt libraries (system / user prompts,

chainofthought, compression).

• Launch experimentdesign module (fractionalfactorial).

• Mentor incoming ML/Data hires; set MLOps standards.

You might be a fit if you have

5 + yrs

RL or bandit systems

scale.

BigQuery / Snowflake SQL

Pub/Sub).

Handson LLM finetuning using LoRA/QLoRA

promptengineering skills

compression).

GCP

baremetal K8s).

 Solid causalinference experience (CUPED, diffindiff, synthetic control, uplift).

 CI/CD, IaC (Terraform or Pulumi) & observability chops (Prometheus, Grafana).

 Bias toward shipping working software over polishing research papers.

Bonus points for:

 Postal/geo datasets, adtech, or martech domain exposure.

 Packaging RL models as secure microservices.

•  VPCSC, NIST, or SOC2 controls in a regulated data environment.