Insight And Analytics Lead (Pharma)
KAI Conversations ·
KAI Conversations is an AI-powered conversation analytics platform used by some of the world's largest pharmaceutical companies to improve the quality and effectiveness of their customer interactions. The KAI platform supports multiple languages, integrates with enterprise systems, and enables large organisations to understand and improve the impact of their customer conversations.
We turn conversation analytics and CRM data into insight that pharma brand teams, local operating companies, field sales managers, and Reps/MSLs can act on. We're looking for an analyst who can take on almost any BI or insight challenge a client raises — a new dataset, a new audience, a new question - and work it through end to end, from raw data to a report that changes what someone does next.
The hardest calls here aren't statistical, they're judgment calls: knowing when a finding is clinically or commercially plausible rather than just tidy-looking, understanding how field force, marketing, and market access functions actually operate, and knowing what “good” looks like in a pharma commercial context.
That judgment is what separates a correct-looking chart from a finding a client can act on. You'll use modern AI tools (Claude, Copilot, ChatGPT) as core working infrastructure for analysis and reporting - and help make that work reliably and repeatably, so good analysis isn't rebuilt from scratch each time.
That's one capability among several: you'll flex across whatever data sources, stakeholders, and formats a client challenge calls for. As this work matures into repeatable, automated outputs, you'll also help shape dashboard concepts for different stakeholders - brand marketing, LOCs, field sales, Reps/MSLs - working with our UI/UX team to bring them to life.
No design tool experience (e.G. Figma) is needed; what matters is knowingwhat each stakeholder needs to see, and communicating that clearly to the people who build it. What kinds of questions you will be answering? The specifics vary by client and brand, but here are some examples: Which objections are actually blocking adoption of a therapy, and how well are reps resolving them in the room?
What do patients themselves raise as concerns or barriers, and where does that create friction in the care pathway? What separates a brand's best-performing reps from the rest, in terms of what they actually do differently?
How would linking CRM and prescribing data change what we could confidently tell a brand team next quarter? You should expect to move between questions like these regularly, often for different brands and different stakeholder audiences in the same week.
Key Responsibilities Analysis, across a genuinely broad remit Combining conversation analytics with CRM data (Veeva, Salesforce, and others) to answer questions neither source can answer alone Auditing new datasets end to end, flagging data quality issues before drawing conclusions Adapting to whatever BI challenge a client brings — new data, new questions, new formats — rather than a fixed playbook Using AI tools (Claude, Copilot, ChatGPT) to explore datasets efficiently, while remaining the check on whether a finding actually holds up Judgment and rigor Sense-checking findings against real commercial and clinical context — treatment pathways, competitor dynamics, field operations, compliance — not just the data itself Reconciling disagreements between metrics or data sources by understanding what each actually captures Saying when a finding doesn't hold up, and why, rather than presenting something tidy but hollow Reporting, storytelling, and stakeholder work Producing strong, infographic-style visuals designed to make a specific finding land clearly, not default charts Reframing the same findings into different reports for different audiences — brand marketing, LOCs, field sales, Reps/MSLs — without distorting the evidence Presenting findings to stakeholders as part of a team, defending methodology under challenge Building reliable, repeatable capability Turning proven analysis approaches into structured, documented, reusable methods, reducing rebuild effort for each new client or dataset Owning prompt design and refinement for AI-driven analysis, treating reliability as engineered, not assumed Documenting known pitfalls and methodology decisions as they're discovered Product