About the role
You own the data and the truth about how our agents perform. You turn raw call data into reports, dashboards, and evaluations that clients and the internal team rely on — and you run the evals that tell us whether an agent is genuinely good, not just live. When a client asks "is this working?", your numbers answer the question.
What you'll own
Data & reporting
- Pull, clean, and structure call and agent data across platforms.
- Build and maintain the recurring reports clients receive — weekly QA summaries and performance reports.
- Own data accuracy and definitions: what counts as containment, resolution, escalation, and so on, applied consistently across every account.
Analytics & dashboards
- Build and maintain dashboards — both internal and client-facing — covering agent performance, call volumes, and trends.
- Surface the why, not just the what: where containment dropped, which intents misfire, where agents fail and why.
Evaluations (evals)
- Design and run evaluations on agent calls and outputs, scored against each agent's approved scope rather than a generic standard.
- Score transcripts, run weekly QA batches, and quantify blockers so they can be prioritized.
- Produce the client assurance summary that goes out with each reporting cycle.
Performance ownership
- Own the measurement-side picture of agent performance and close the loop with the Conversational AI Engineer: you find the problem, they fix the agent, you confirm the fix landed.
- Maintain the QA workbook and per-agent evaluation packs.
What a strong first 90 days looks like
- You can independently produce a client-ready weekly performance and QA report.
- You've stood up or taken over at least one live dashboard.
- You can run an eval batch on a new agent and explain, in plain language, where it's strong and where it's failing.
Required
- Strong data skills: SQL, advanced spreadsheets (Excel or Google Sheets), and comfort manipulating and reconciling data.
- Hands-on dashboard / BI experience (Looker, Power BI, Metabase, or similar).
- Analytical rigor — you care about definitions and accuracy, and you don't ship numbers you can't defend.
- A clear communicator who can write a summary a client trusts.
Nice to have
- Python (pandas) for data work.
- Experience evaluating LLM or AI outputs.
- Familiarity with conversation analytics or contact-center reporting.
- Exposure to Postgres and data pipelines.
Tools you'll work with
SQL / Postgres · Excel / Google Sheets · a BI tool · Python · VerSight and the QA workbook · call-platform data exports.
Reporting line
Reports to the BI / Data & Delivery lead.
Fill below form to apply: (Only form filled applications will be considered)
https://docs.google.com/forms/d/e/1FAIpQLScdV05XKLt4-OvJq8j1HCdZlHdNxIfHoEBQeuesuTzBqapOlg/viewform?usp=publish-editor
Pay: ₹800,000.00 - ₹1,000,000.00 per year
Work Location: Remote
