Case Study
Dala.AI — Designing the Conversation Intelligence Platform
End-to-end product design for Dala.AI, a conversation intelligence platform that transforms raw sales call data into structured insights, coaching opportunities, and deal-risk signals.
Overview
Dala.AI is a conversation intelligence tool built for revenue teams. It records, transcribes, and analyses sales calls—surfacing deal risks, objection patterns, and coaching moments in real time. The design challenge was to make sense of dense, multi-layered call data without overwhelming reps or managers, and to surface the right signal at the right moment in the workflow.
Why This Product
Sales managers were spending hours manually reviewing call recordings to identify coaching moments, track competitor mentions, and assess deal health. Reps had no structured feedback loop between calls. The existing tools provided transcripts but no actionable synthesis— leaving teams to derive insight through manual effort.
How might we surface the most critical call insights to both reps and managers without requiring them to watch hours of recordings or parse raw transcripts?
Users
The platform served two primary personas with distinct workflows and goals. Sales reps needed immediate, actionable feedback after each call. Managers needed a macro view across the team—identifying patterns, flagging at-risk deals, and running targeted coaching sessions.
Sales Reps
Need post-call summaries, talk-track feedback, and clear next steps to improve performance between conversations.
Sales Managers
Need a team-level view of call quality, deal risk signals, and coaching opportunities without reviewing every recording.
My Approach
I began with stakeholder interviews and shadowing live sales calls to understand how reps and managers actually consumed call data today. From there I mapped the post-call workflow, ran competitive analysis, and iterated through low-fi wireframes before moving into high-fidelity prototypes tested with real users.
- Stakeholder Interviews
- Call Shadowing
- Workflow Mapping
- Competitive Analysis
- Wireframes
- Usability Testing
- High-fi Prototypes
- Handoff
Existing Workflow & Pain Points
Before Dala.AI, reps relied on manual note-taking during calls and CRM updates done from memory afterward. Managers had no visibility into call quality until deal outcomes were already affected. Key pain points were mapped across the entire call lifecycle—pre-call prep, live call, and post-call review.
Research
Interviews with 18 reps and 8 managers surfaced recurring themes. We scored each pain point by frequency and severity to prioritise the design focus areas.
Post-Call Notes
Coaching Visibility
Deal Risk Detection
Competitor Mentions
Talk-Listen Ratio
Next-Step Clarity
User Flows
Rep Post-Call Flow
Pain points clustered around the gap between call end and CRM entry: reps forgot key details, next steps went unlogged, and feedback arrived too late to act on.
Redesigned Post-Call Flow
The redesigned flow automated summary generation, surfaced next steps inline, and pushed structured data to the CRM — reducing manual effort from ~12 minutes to under 2.
Final Design
Call Summary View
Each call surfaces a structured summary with key moments, objections raised, and committed next steps — all timestamped and linked back to the transcript.
Manager Dashboard
- Team-level call quality scores at a glance.
- Deal risk flags surfaced without watching recordings.
- One-click drill-down into specific calls for coaching.
Coaching Moment Flow
- Managers can bookmark key moments and leave timestamped comments.
- Reps receive coaching notes directly within the call summary view.
- Feedback loop closes within the same tool — no context switching.
- Coaching history is tracked over time per rep.
Real-Time Call Assist
- Battle cards surface automatically when competitor names are detected.
- Objection responses are suggested based on what's being said live.
- Talk-listen ratio tracked in real time with a subtle nudge when imbalanced.
Impact
- Reduced post-call admin time from ~12 minutes to under 2 minutes per call.
- Managers identified at-risk deals an average of 4 days earlier.
- New rep ramp time reduced by 30% through structured call feedback.
- Coaching session prep time cut significantly for front-line managers.
- Deal win rates improved in teams with high platform adoption.