AskAxon

Problem

Most companies do not have a data problem. They have an access problem.

The answers exist somewhere - in a warehouse, a BI dashboard, a spreadsheet, or a report someone built six months ago - but the people who need those answers usually have to wait. They ask an analyst, search through dashboards, or make a decision with partial information because the cost of finding the answer is too high.

At Certn, I kept seeing the same pattern: smart people with good business questions, blocked by the friction of getting to the data. The opportunity was not to replace BI. It was to create a lightweight layer on top of it - something fast enough for everyday questions and simple enough that non-technical teams would actually use it.

Approach

I built AskAxon as a Slack-native AI data assistant.

The first version was intentionally lightweight. It connected to existing data sources and BI infrastructure, then allowed employees to ask questions in plain English from the place they already worked: Slack.

The product translated business questions into data queries, returned answers in natural language, and helped users understand the context behind the number instead of just showing them another chart. The goal was not to build a full analytics platform. It was to shorten the distance between a question and a useful answer.

I focused heavily on workflow fit. AskAxon did not ask users to learn a new dashboarding tool or change how they worked. It met them where the question already happened.

Outcome

AskAxon was adopted across 200+ employees at Certn and later used by BPD Healthcare's marketing team.

More importantly, it validated the core insight behind my AI work: people do not just want access to data. They want the confidence to make decisions with it.

That feedback eventually became the foundation for DataSays - a more powerful standalone AI data workspace designed to move beyond simple Q&A into insights, charts, dashboards, slides, and narrative outputs.

Stack

• OpenAI - natural language understanding and answer generation
• Slack API - conversational user interface
• Python / FastAPI - backend application layer
• LangChain - orchestration and tool use
• SQL / BI connectors - data access layer
• Data warehouse / BI layer - source-of-truth infrastructure