AI Voice Agent Optimization Specialist
Company Description
Fostr connects growing businesses with highly skilled freelancers from global talent pools, focusing on quality and long-term collaboration. The company leverages a trusted network of professionals to match client needs with the right expertise. This approach emphasizes reliability, transparency, and a high standard of care in every engagement. By combining global reach with a curated network, Fostr offers flexible, remote-friendly opportunities for specialists who want to work with diverse clients and industries.
Client Description.
The client runs performance marketing campaigns in the call space — taking raw data, building audiences, and deploying them into an agentic outbound call platform. The client uses
TAALK which is an agentic call platform where the agents and settings are configured.
This individual would manipulate scripts that are automatically dialing to prospective clients to gauge interest for a product - once the consumer is qualified we live transfer them to a person to close the deal.
Hours : 9am-5pm EST Type : Contractor Rate : $10USD / hour (8 hours per day)
Role Overview
We are seeking a hands-on Agentic Performance & Optimization Specialist to improve the performance, reliability, and compliance of our outbound AI voice agents. This person will own the day-to-day evaluation and optimization of live agent interactions, including script testing, prompt refinement, call-quality analysis, operational troubleshooting, and compliance auditing. This is an individual-contributor role. The ideal candidate is analytical, technically comfortable, and able to translate call data and customer interactions into measurable improvements in billable transfer rate, call duration, stability, and overall customer experience.
Key Responsibilities
Script Optimization and Variant Testing
Identify call drop-off points, objection patterns, and areas of friction within live scripts. Develop “challenger” prompt and script variants that test changes to pacing, tone, qualification, objection handling, and call flow. Stress-test new variants in a sandbox environment before introducing them to live traffic. Validate prompt stability, guardrails, API calls, function calling, and transfer workflows. Allocate a controlled percentage of traffic to new variants and measure results against the current “champion.” Analyze performance using key metrics such as billable transfer rate, call duration, conversion rate, and failure rate. Recommend and implement winning variants in production.
Daily AI Agent Operations
Serve as the first point of investigation for daily agent-performance issues, including latency, speech-to-text errors, transfer failures, and unexpected agent behavior. Review logs and call events to ensure transfers, handoffs, voicemail handling, and call-screening workflows are functioning correctly. Apply prompt-level corrections when appropriate and escalate technical or integration issues to the appropriate team. Provide concise reports summarizing issues identified, corrective actions taken, and recommended next steps. Maintain organized testing records, change logs, and performance documentation.
Call Quality and Interaction Analysis
Review a daily sample of dropped, failed, and underperforming calls. Categorize root causes, including agent behavior, poor lead data, call timing, technical failures, and buyer-side issues. Identify new or poorly handled customer objections and recommend prompt or script adjustments. Improve voice quality and human-likeness by adjusting pauses, pacing, pronunciation, emphasis, and conversational language. Track recurring call patterns and convert findings into actionable improvements.
Compliance and Guardrail Testing
Red-team AI agents to confirm they appropriately handle out-of-scope questions and avoid providing unlicensed financial, insurance, or medical advice. Audit transcripts for required disclosures, approved language, and explicit prospect consent. Identify and document compliance failures or risky agent behavior. Test and refine guardrails, escalation paths, fallback responses, and disclosure logic.
Required Qualifications
Experience with conversational AI, voice AI, prompt engineering, call-center technology, performance operations, or a related field. Hands-on experience writing, testing, and optimizing prompts or call scripts using live performance data. Demonstrated ability to leverage AI tools—such as Claude, ChatGPT, or similar platforms—to analyze interactions, generate hypotheses, improve prompts, and accelerate testing. Experience with A/B testing or champion/challenger testing methodologies. Ability to analyze call transcripts, performance funnels, dashboards, and operational logs. Understanding of outbound calling workflows, including transfers, voicemail handling, call screening, lead routing, and buyer integrations. Familiarity with speech-to-text, text-to-speech, APIs, webhooks, function calling, and system integrations. Strong understanding of metrics such as billable transfer rate, conversion rate, call duration, connection rate, and failure rate. Strong written communication and documentation skills. Ability to work independently, prioritize issues, and respond calmly to urgent production problems.
Fostr connects growing businesses with highly skilled freelancers from global talent pools, focusing on quality and long-term collaboration. The company leverages a trusted network of professionals to match client needs with the right expertise. This approach emphasizes reliability, transparency, and a high standard of care in every engagement. By combining global reach with a curated network, Fostr offers flexible, remote-friendly opportunities for specialists who want to work with diverse clients and industries.
Client Description.
The client runs performance marketing campaigns in the call space — taking raw data, building audiences, and deploying them into an agentic outbound call platform. The client uses
TAALK which is an agentic call platform where the agents and settings are configured.
This individual would manipulate scripts that are automatically dialing to prospective clients to gauge interest for a product - once the consumer is qualified we live transfer them to a person to close the deal.
Hours : 9am-5pm EST Type : Contractor Rate : $10USD / hour (8 hours per day)
Role Overview
We are seeking a hands-on Agentic Performance & Optimization Specialist to improve the performance, reliability, and compliance of our outbound AI voice agents. This person will own the day-to-day evaluation and optimization of live agent interactions, including script testing, prompt refinement, call-quality analysis, operational troubleshooting, and compliance auditing. This is an individual-contributor role. The ideal candidate is analytical, technically comfortable, and able to translate call data and customer interactions into measurable improvements in billable transfer rate, call duration, stability, and overall customer experience.
Key Responsibilities
Script Optimization and Variant Testing
Identify call drop-off points, objection patterns, and areas of friction within live scripts. Develop “challenger” prompt and script variants that test changes to pacing, tone, qualification, objection handling, and call flow. Stress-test new variants in a sandbox environment before introducing them to live traffic. Validate prompt stability, guardrails, API calls, function calling, and transfer workflows. Allocate a controlled percentage of traffic to new variants and measure results against the current “champion.” Analyze performance using key metrics such as billable transfer rate, call duration, conversion rate, and failure rate. Recommend and implement winning variants in production.
Daily AI Agent Operations
Serve as the first point of investigation for daily agent-performance issues, including latency, speech-to-text errors, transfer failures, and unexpected agent behavior. Review logs and call events to ensure transfers, handoffs, voicemail handling, and call-screening workflows are functioning correctly. Apply prompt-level corrections when appropriate and escalate technical or integration issues to the appropriate team. Provide concise reports summarizing issues identified, corrective actions taken, and recommended next steps. Maintain organized testing records, change logs, and performance documentation.
Call Quality and Interaction Analysis
Review a daily sample of dropped, failed, and underperforming calls. Categorize root causes, including agent behavior, poor lead data, call timing, technical failures, and buyer-side issues. Identify new or poorly handled customer objections and recommend prompt or script adjustments. Improve voice quality and human-likeness by adjusting pauses, pacing, pronunciation, emphasis, and conversational language. Track recurring call patterns and convert findings into actionable improvements.
Compliance and Guardrail Testing
Red-team AI agents to confirm they appropriately handle out-of-scope questions and avoid providing unlicensed financial, insurance, or medical advice. Audit transcripts for required disclosures, approved language, and explicit prospect consent. Identify and document compliance failures or risky agent behavior. Test and refine guardrails, escalation paths, fallback responses, and disclosure logic.
Required Qualifications
Experience with conversational AI, voice AI, prompt engineering, call-center technology, performance operations, or a related field. Hands-on experience writing, testing, and optimizing prompts or call scripts using live performance data. Demonstrated ability to leverage AI tools—such as Claude, ChatGPT, or similar platforms—to analyze interactions, generate hypotheses, improve prompts, and accelerate testing. Experience with A/B testing or champion/challenger testing methodologies. Ability to analyze call transcripts, performance funnels, dashboards, and operational logs. Understanding of outbound calling workflows, including transfers, voicemail handling, call screening, lead routing, and buyer integrations. Familiarity with speech-to-text, text-to-speech, APIs, webhooks, function calling, and system integrations. Strong understanding of metrics such as billable transfer rate, conversion rate, call duration, connection rate, and failure rate. Strong written communication and documentation skills. Ability to work independently, prioritize issues, and respond calmly to urgent production problems.