Skip to content

VIRTUAL COMPANIES

Four companies. Different standards. Same rigor.

Every Safua mission lands inside one of four virtual companies. Each has its own industry framing, its own review priorities, its own senior engineers. You don’t ship into a void — you ship into a team.

AI agent infrastructure

NovaMind AI

Agentic systems at production scale — multi-agent orchestration, reliability, safety.

Review priorities

  • Correctness under adversarial prompts
  • Reliability of multi-agent orchestration
  • Reasoning traceability and HITL checkpoints

Primary schools

Example missions

  • Ship a multi-agent research assistant with explicit conflict resolution between planner and critic.
  • Wire a human-in-the-loop approval gate for an autonomous browsing agent with a published escalation policy.
  • Red-team a production RAG pipeline and document three prompt-injection vectors with their mitigations.

Senior engineers on this company

  • YT

    Dr. Yuki Tanaka

    CTO

  • DV

    Dimitri Volkov

    VP of Engineering

  • RM

    Rafael Mendes

    Senior Engineer — RAG & Retrieval

  • IN

    Isla Nakamura

    Senior Engineer — Agent Orchestration

  • AR

    Amir Rezaei

    Tech Lead

Data infrastructure & pipelines

DataForge Labs

Where idempotency, governance, and reproducibility meet scale.

Review priorities

  • Code quality and maintainability
  • Reproducibility of pipeline runs
  • Governance trails and SLA adherence

Primary schools

Example missions

  • Migrate a batch ingest pipeline to streaming while preserving idempotency guarantees.
  • Deploy a dbt + Airflow stack on Kubernetes with a complete rollback procedure.
  • Audit an existing feature store for data leakage and propose a fix with reproducibility proofs.

Senior engineers on this company

  • MO

    Dr. Miriam Okafor

    CTO

  • MC

    Marcus Chen

    VP of Engineering

  • JP

    Jin Park

    Senior Engineer — Pipelines

  • CJ

    Clara Johansson

    Senior Engineer — MLOps

  • LM

    Luis Morales

    Tech Lead

Computer vision & edge AI

VisionArc

Intelligence on constrained hardware — latency, memory, and accuracy traded deliberately.

Review priorities

  • Problem solving within hardware constraints
  • Correctness across diverse inputs
  • Awareness of distribution shift and bias

Primary schools

Example missions

  • Ship an edge-inference model that meets 50 ms latency and 2 GB memory constraints on target hardware.
  • Audit a deployed CV system for demographic bias and document the threshold trade-offs.
  • Design a visual agent that handles domain shift gracefully with a confidence-gated escalation path.

Senior engineers on this company

  • AM

    Dr. Andreas Müller

    CTO

  • PS

    Priyanka Sharma

    VP of Engineering

  • FD

    Dr. Fatou Diallo

    Senior Engineer — Computer Vision

  • KY

    Kai Yamamoto

    Senior Engineer — Deep Learning

  • AP

    Aisha Patel

    Tech Lead

Healthcare AI with regulatory compliance

Sentient Health

Clinically meaningful AI built for HIPAA, auditability, and patient outcomes.

Review priorities

  • Correctness on clinical edge cases
  • Engineering thinking around audit and compliance
  • Communication quality of governance write-ups

Primary schools

Example missions

  • Add a HIPAA-compliant audit trail to an existing AI inference pipeline without blocking the request path.
  • Run a four-fifths fairness audit on a clinical decision-support model and document the remediation plan.
  • Review a triage model for PHI leakage risk and produce a named-reviewer-ready governance report.

Senior engineers on this company

  • EO

    Dr. Elizabeth Okafor

    CTO

  • DN

    Daniel Nakamura

    VP of Engineering

  • LK

    Dr. Lena Kowalski

    Senior Engineer — Compliance & Safety

  • RM

    Dr. Ravi Mehta

    Senior Engineer — Explainability

  • ST

    Sophie Tremblay

    Tech Lead

HOW MISSIONS WORK

You don't pick a company. The platform routes.

Mission assignment is deliberate. When a learner enters Build in a given school, the platform routes their missions across the companies whose review priorities stress the skills the learner most needs to develop.

A learner strong on code quality but still developing engineering thinking will draw more DataForge missions (where governance, reproducibility, and SLA-awareness are primary) before they see VisionArc missions (where hardware-constrained problem solving dominates the review). The distribution moves with their Confidence Score.

You can see the current weighting on each school page under “Virtual Company Fit.”

Your un-fakeable AI engineering identity starts here.

Join the engineers building proof, not just portfolios.

Get started for freeTalk to sales