SHRIYA MOTHILALL
Chat Studio
Can your words turn frustration into loyalty?
Flagship Project: Chat Studio
This live-chat simulation places customer service agents in realistic digital conversations where they must navigate tone, empathy, response timing, and problem-solving under pressure. Learners practice handling different customer personalities and scenarios, receive instant feedback, and see how their choices affect customer satisfaction ratings and overall experience.
.png)
THE PROBLEM
In high-volume call centers, agents are expected to deliver fast, empathetic, brand-consistent service through live chats. ​ However, agents have been receiving lower customer satisfaction scores - directly affecting brand perception and operational efficiency at scale. Most training relied on static modules or scripts that don’t reflect real-world pressure or unpredictable customer behavior and objections. Due to this, agents are not getting the effective learning experience before going live and handling real customers.
​
Fixing the issue isn’t simple: strong chat performance requires emotional intelligence, systems fluency, and split-second judgment, while individualized coaching is costly and difficult to scale across large teams.​ Without a safe way to practice realistic scenarios, agents are left to learn on live customers - creating risk for the business and long ramp-up times for new hires.​
THE SOLUTION
This project delivers a live-chat simulation experience that lets agents practice realistic customer conversations before going live. Through a series of eLearn-modules, learners respond to different customer personalities, emotional states, and service challenges while managing tone, pacing, and clarity in real time.
The project is built to "live-rate" the agents interactions, providing real time correction and thus an effective learning experience. Instant customer-rating outcomes show how each response impacts satisfaction, escalation risk, and coming to an overall resolution.
Integrated into the LMS, the simulation supports repeat practice, targeted coaching, and data-driven insights for leaders - creating a scalable way to build confidence, standardize service quality, and shorten onboarding time across teams.
AUDIENCE
Customer Service Agents
RESPONSIBILITIES
-
Content Writing
Scripts, storylines, learner prompts
​
-
Instructional Design
Action mapping, storyboarding)
​
-
eLearning Development
Prototyping, full build, visual design
TOOLS USED
-
Articulate Storyline 360
-
Adobe Illustrator
-
FreePik
-
Canva
-
Google Docs
-
Pixabay
My Process
I followed the ADDIE model to guide the project, beginning with action mapping in collaboration with an experienced customer service agent, who served as the SME. Using insights from our meetings, I created a text-based storyboard, visual mockups in Illustrator, and a Storyline prototype, which was refined based on the SME’s feedback before full development.
Action Map
After discussing the problem and proposed solution, I consulted with my SME to identify actions that current and new agents were performing incorrectly. Action mapping helped is focus on what agents need to do – not just what they need to know – by zeroing in on observable, high-impact behaviours and techniques that directly support the business goal of improving customer satisfaction, and brand reputation. Together, we prioritized key principles that should be applied within each interaction in Chat Studio, including first impression techniques, empathising and professionalism and correct use of tone. We included the current go-to behaviours from agents alongside the correct behaviours, to allow them compare, self-reflect and correct themselves.
Text-based Storyboard
Next, I worked on the text-based storyboard with my SME. This was a vital step, serving as the blueprint for the project design and development. It was written so that each interaction required the application of customer service techniques and principles that users have already learnt. Collaborating with my SME ensured all messages and responses modelled real-life as closely as possible, immersing the user in a realistic interaction that they may come across. As guidance to the learner, I introduced a mentor chat that assists the user with navigation controls, the objective of the module, as well as information and tips to consider. While we considered using mentor support for each interaction, we concluded it would be more effective to allow agents to make decisions in real-time, without extra support. I also created a customer satisfaction rating bar using stars - a gaming component to help encourage the learner and indicate behaviours that could hinder customer satisfaction. Stars increase if the correct response is chosen, and decreases for incorrect responses. The end message congratulates high-performers, and encourages those who struggled by suggesting additional support will be provided to avoid discouragement.
Prototyping
I moved into prototyping the eLearning experience using Articulate Storyline 360. The interactive prototype included the title screen, mentor introduction, and the 1st - 3rd interactions correct and incorrect consequences. I focused on capturing the UX/UI to ensure the navigation felt natural and the experience remained immersive.
I prototyped the chat movement to give a scrolling effect, where animations were synchronized with audio to replicate text message notifications. Additionally, I introduced the first version of the customer ratings bar - a visual tool designed to reflect the consequences of learner choices in real time.
The prototype was shared with key stakeholders, including existing customer service agents, training instructors and my content development team to gather feedback on usability, flow, and visual design. Based on the feedback, I made minor refinements, including tightening transition timings and Storyline, and adding audio cues. With the core framework validated, I completed development and launched the project successfully.
Measuring Results & Takeaways
REACTION
Using a Google Form survey we gathered learner feedback on their course experience and their perceived readiness for real-world scenarios.
LEARNING
I then created a rubric aligned with the six key actions identified in the action map. This rubric will allow customer service staff to assess whether learners gained the intended skills as well as providing the targeted feedback to support continuous improvement.
BEHAVIOUR
The next step of evaluation included analysing responses to each key action, with the goal being 80% of observed agents to demonstrate proficiency in at least 4 out of 6 targeted actions – showing the ability to apply their theoretical knowledge.
RESULTS
Incident data was tracked monthly for 6 months between 2 groups: Group A (untrained agents) and Group B (trained agents). Baseline data from the previous 6 months will serve as a reference point. Staff will log incidents in a shared tracking document, and at the end of the 6-month period, we will analyse the data to determine whether the training resulted in a measurable reduction in incidents.