Ather Energy's support agents were manually summarising every call — under time pressure, at scale, with no consistency. BootLabs deployed a GenAI agent on Google Cloud that automated the entire process.
Ather Energy's customer support team handles a significant volume of inbound calls daily — spanning technical inquiries about the Ather 450 EV scooter, financial queries, service scheduling, and escalations. After each call, agents were required to manually summarise the interaction during a short post-call disposition window. Under time pressure and at scale, the quality of these summaries dropped significantly — key details were omitted, action items went unrecorded, and patterns across thousands of customer conversations were never captured. The intelligence locked inside those calls — product feedback, recurring service issues, escalation signals — was invisible to the product and operations teams who needed it most. As Ather Energy's Google Cloud partner, BootLabs designed and deployed a GenAI-powered call summarisation agent that automated this process entirely, turning every call into structured, CRM-ready data.
Agents had to summarise lengthy, multi-topic calls within tight post-call windows. Technical inquiries, financial queries, service scheduling, and escalations often overlapped within a single call. Under time pressure, quality dropped — key details were omitted, action items went unrecorded, and the next call was already waiting.
Call content was never systematically captured for analysis. Patterns in customer complaints, feature requests, and service failures were invisible — preventing product and ops teams from acting on voice-of-customer data. Thousands of high-signal conversations were disappearing into unstructured notes with no downstream value.
Summary quality varied widely across the support team — some agents wrote thorough notes, others wrote one-liners. No standardisation meant CRM records were unreliable as a data source. Downstream analysis, escalation tracking, and reporting were all compromised by this inconsistency.
Google Cloud's Speech-to-Text API transcribes every customer-agent call in real-time. The pipeline is tuned for Indian English accents, domain-specific EV terminology across the Ather 450 product line, and multi-speaker audio — delivering high-accuracy transcripts that form the foundation for everything downstream.
A GenAI model (Google Gemini / PaLM) processes each transcript and produces a structured call summary: key topics discussed, customer sentiment (positive / neutral / frustrated / escalation risk), action items, and follow-up owners. Engineered prompts ensure output is consistent, structured, and CRM-ready — every time, regardless of call complexity.
Summaries, sentiment tags, and action items are pushed directly into the relevant Salesforce case record via API integration — with zero agent effort required. The record is fully updated before the next call begins, eliminating the post-call disposition window entirely and freeing agents immediately for their next interaction.
Structured call data accumulates into a customer intelligence layer: trending issues, sentiment by product line, escalation patterns, and repeat contact reasons. This data surfaces to product and operations leadership as actionable dashboards — giving Ather Energy a structured voice-of-customer feed for the first time.
Every call Ather Energy's support team takes is now automatically transcribed, summarised, sentiment-tagged, and pushed into Salesforce — before the agent picks up the next call. The manual disposition step has been eliminated entirely. For the first time, Ather Energy has a structured, searchable record of every customer interaction, with consistent quality across the entire support organisation.
Post-call disposition is eliminated entirely — agents no longer spend time writing summaries under pressure. Cognitive load drops, and agents move to their next interaction faster with full focus on the customer rather than the keyboard.
Product and ops teams at Ather Energy now have structured voice-of-customer data flowing into analytics for the first time. Recurring service issues, feature requests, and sentiment trends across the 450 series lineup are visible and actionable.
Summary quality is no longer dependent on individual agent effort or attention. Every Salesforce record is structured identically — enabling reliable downstream analysis, escalation tracking, and reporting that the team couldn't trust before.
The structured call data layer creates the foundation for the next generation of AI-driven capabilities at Ather Energy: predictive escalation flagging, proactive service outreach based on sentiment trends, and intelligent routing informed by call history.
Tell us about your challenge and we'll set up a focused 30-minute session.