Skip to main content
Back to challenges
Challenge

Voice-Enabled Tanker Allocation Agent using OpenAI Agents SDK and Retell AI

When managing immediate short-term crude exports across Basra terminals, operations teams require rapid voice-driven updates and automated dispatch actions. Build a voice-interactive tanker allocation agent that leverages Retell AI for conversational interaction and OpenAI Agents SDK for background execution and tool calling. The agent must parse real-time voice queries from terminal dispatchers, check berth schedules, verify 3-month temporary export permits, and automatically queue vessel loading sequences.

Business OperationsHosted by Vera
Sign in free to start
Challenge brief

What you are building

The core problem, expected build, and operating context for this challenge.

Develop a real-time voice and agent dispatch system for oil tanker loading schedules using OpenAI Agents SDK and Retell AI.

Delivery guide

How work is evaluated

Evaluation

Tests voice transcript intent mapping and accurate berth assignment generation.

Datasets

Challenge datasets

Public datasets, plus the private datasets you have access to.

Evaluation rubric

How submissions are scored

These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.

Dimensions
2 scoring checks
Binary
2 pass or fail dimensions
Ordinal
0 scaled dimensions
Dimension 1

verify_berth_assignment

Ensure requested berth is allocated accurately.

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Dimension 2

intent_accuracy

Accuracy of intent extraction from spoken input. • target: 0.98 • range: 0-1

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Learning goals

What you should walk away with

  • Implement real-time voice agent workflows using Retell AI integration.

  • Build function-calling agents with OpenAI Agents SDK for tanker schedule updates.

  • Synchronize voice intent parsing with transactional database modifications.

Resources and assets

Reference links and supporting material

Dataset notes

Audio-transcript pair dataset of terminal dispatcher commands and terminal port states.

Challenge setup

Tests voice transcript intent mapping and accurate berth assignment generation.

Challenge input

JSON with transcript and current_terminal_state

OpenAI Agents SDK

Lightweight autonomous agent execution engine.

Retell AI

Provides real-time voice streaming integration.

Evaluated output

JSON with target_vessel, assigned_berth, action_status

Evaluation checks
  • Ensure requested berth is allocated accurately.
  • Accuracy of intent extraction from spoken input. • target: 0.98 • range: 0-1
Configured evaluation evidence
  • Benchmark: Voice-Logistics-Bench
  • Intent Accuracy target: 0.98
  • 1 public reference case
Configured runtime
  • Python execution harness
  • Python sandbox (unavailable on Versalist)

Summary shows the first task, up to three tools, and up to three evidence items. Open the technical recipe to view all configured items.

View technical recipe

Configured tools

Action Space
  • Retell AI · Required
  • OpenAI · Optional
  • MLflow · Optional

Evaluation contract

  • verify_berth_assignment · Weight 1
  • intent_accuracy · Weight 1

Metrics

  • intent_accuracy · Target 0.98

Benchmarks

  • Voice-Logistics-Bench
  • general

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON with transcript and current_terminal_state

    Output: JSON with target_vessel, assigned_berth, action_status

    Execution harness: Python

Evaluation tools

  • OpenAI Agents SDK · Agent Orchestration
  • Retell AI · Voice Gateway

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Voice-Enabled Tanker Allocation Agent using OpenAI Agents SDK and Retell AI