A new agent starts on Monday.
You’ve got a backlog of unresolved tickets, a customer satisfaction dip you can’t fully explain, and three senior reps who’ve flagged burnout. You know they need help. But onboarding takes weeks. QA can’t catch everything. And by the time you spot a pattern, it’s already cost you in customer trust.
Every support leader wants the same thing: a team of human agents who can handle complex issues with speed, empathy, and consistency. But older tools weren’t designed to make that easy. They offered static suggestions, generic macros, or lagging QA reports that arrive long after the moment has passed.
We believe their AI-powered successors should work like teammates, so we updated our Assist Agent with agentic AI and developed the brand-new Agent QA. Together, they understand what each agent is dealing with, guide them through complex workflows, and help them respond with clarity and confidence while improving over time.
Support teams need tools that elevate agents
Most tools that claim to be AI-powered promise efficiency but deliver rigidity. They suggest prewritten templated responses, surface loosely related knowledge base articles, or flag tickets for manual review days after the fact. They might shave off a few seconds per response, but they don’t actually help agents think through complex problems or improve over time.
Forethought takes a different approach. We believe that in order to deliver on the promise of agentic AI, AI should assist in the moment and take meaningful action to help agents get better with every ticket.
- Assist Agent uses it to support human agents in real time
- Agent QA builds on that support by measuring what’s working and where to improve.
Together, they create a smarter system that learns from every conversation and helps your team perform at its best. Built for internal teams, they’re designed for the people who feel the pain of inefficient support firsthand: support leaders, QA managers, and the agents themselves.
“Agent Assist makes agents faster in the moment; Agent QA makes them better over time. Working in tandem, the two form a continuous improvement loop: Assist guides each reply, QA scores the outcome, and our platform learns from both—so every ticket is handled smarter than the last.” — Forethought CTO, Jad Chamoun
Think of Assist Agent like your AI copilot, delivering ticket summaries, step-by-step resolution guidance, and accurate responses right in the helpdesk, and Agent QA as your always-on evaluator, helping managers track performance across customizable metrics.
Assist Agent acts like a seasoned teammate inside the browser
Most tools promise to make support agents faster. But in practice, they bury the agent in tabs or force them to copy-paste outdated knowledge base links. Nothing about that setup helps someone solve a complex ticket with confidence.
Assist Agent is a browser-based copilot that works alongside your human agents, right inside the helpdesk. From the moment a ticket opens, it starts working: summarizing the request, analyzing context, retrieving relevant workflows and articles, and drafting a brand-aligned response.
Because it's powered by agentic AI, Assist Agent doesn't just point agents to content. It reasons through the ticket, recommends what to do next, and provides just enough guidance to help the agent move forward quickly. It’s built to behave like a seasoned teammate who knows the playbook and shares what to do right when you need it.
In the example below, the Assist Agent has already summarized the customer request, retrieved a manager-authored workflow, and generated a ready-to-send reply tailored to the issue at hand.

The agent never had to search for context or ask what to do. Assist Agent doesn’t just hand the agent a document and wish them luck. It makes a decision based on the topic, intent, and historical behavior and delivers the next best action.
Maybe an agent is new and handling refund requests for the first time. Assist Agent can pull in the appropriate refund policy and generate a customer-ready response automatically. If the agent encounters a technical error, Assist Agent instantly suggests a step-by-step workflow and links a relevant article.
When an agent isn’t sure how to proceed, they can simply ask. Assist Agent includes a real-time chat experience where agents can describe the situation and receive immediate, guided help.
“With Agent Assist, we engineered an AI copilot that behaves like a seasoned teammate—analyzing every ticket in real time, recommending the exact workflow, and drafting a response that’s ready to send. It’s guidance, not guesswork, delivered right inside the agent’s browser.” — Forethought CTO, Jad Chamoun
This is the type of support that scales—new agents ramp faster, senior agents avoid burnout by skipping repetitive searching, and support leaders get visibility into how Assist Agent is being used with Agent QA.
Agent QA gives managers visibility and coaching opportunities at scale
Ask any support leader how they evaluate agent performance, and the answer usually starts with “We try.” Most teams rely on random ticket sampling, scattered spreadsheets, or a single QA manager trying to review 10,000 conversations a week. It’s not scalable. It’s rarely fair. And it’s never fast enough to drive actual improvement.
Agent QA is a feature within Forethought Discover that automatically evaluates 100% of your support conversations using AI-powered, customizable metrics. You decide what matters, whether that’s resolution quality, technical accuracy, empathy, grammar, or something custom to your needs, and Agent QA does the rest. It connects to your help desk, indexes your historical ticket data, and begins scoring every response across the criteria you’ve defined.
Let’s say you manage a team of 100 support agents. You want a clear view of how well your team is resolving customer issues, so you connect your help desk to Forethought and create a few custom metrics: resolution quality, grammar accuracy, and technical completeness.
Agent QA starts evaluating every ticket automatically. Within hours, you have a full dashboard showing how each agent is performing against those metrics without manual tagging or any spot-checking.

You can also drill into individual agent performance at a glance. Here, we can see a human agent, Alex’s, QA score, resolution rate, and a breakdown across every scoring metric configured.
“Agent QA flips the traditional QA model on its head by scoring 100 percent of support conversations for empathy, accuracy, and resolution quality—no manual sampling required. Teams finally get a transparent, data-driven view of performance and the coaching insights to level-up every agent.” — Forethought CTO, Jad Chamoun
Maybe you’ve already spotted an issue in your dashboard: an agent’s grammar and accuracy scores are consistently underperforming. With Agent QA, you can pull up those tickets, identify patterns, and walk through them together in a 1:1. You’ve now got clear, structured examples to coach from, and more importantly, you can track whether that coaching actually works.
This example radar chart makes it instantly clear where Alex needs coaching: empathy and grammar are trending low. Over the next few weeks, Agent QA continues to evaluate that agent’s responses. If his grammar score improves, you will see the progress. If not, you know it’s time to adjust the approach.

This turns performance evaluation into a feedback loop, not a finger-pointing exercise. Agent QA shows you the trends, measures the impact of coaching over time, and gives every manager a real way to improve support quality at scale.
Together, they work better together to improve every conversation
These tools were designed for the teams behind the scenes—support agents, managers, and QA leaders who need real solutions in real time. Together, they help support teams move faster, train smarter, and deliver higher-quality responses across every channel.
Both are available as add-ons for customers of Forethought, making it easy for our customers to extend what they already use. Talk to your team today about adding Assist Agent and Agent QA, and see how much more your agents can do with the right support.
Hashtags blocks for sticky navbar (visible only for admin)
{{resource-cta}}
{{resource-cta-horizontal}}
{{authors-one-in-row}}
{{authors-two-in-row}}
{{download-the-report}}
{{cs-card}}

