Scope is a hard boundary
Every connector defaults to scope=nothing. You opt in to specific tables and fields — anything outside that boundary is invisible to the LLM, the planner, and the executor.
About
AnalystIQ turns the question-to-answer loop from days into seconds — without making your team trust a black box. Built by Algoscale, currently in private beta.
Why we built this
The people with questions and the people who can answer them are never the same. Operators wait days for a number; analysts spend their week translating Slack messages into SQL instead of doing the work they were hired to do.
Existing BI tools punt the problem upstream. Dashboards take months of committee work to ship, then go stale the moment the business changes shape. The good questions — the ones nobody anticipated — never get asked because the cost of asking is too high.
AnalystIQ collapses that loop. Connect a source, curate the semantic layer once, then ask in plain English. The query plan ships with every answer so the analysts on your team stay in the loop — auditing, refining, and trusting what comes back.
How it works
A multi-agent pipeline — schema selector, planner powered by Claude Opus 4.7 with adaptive thinking, a sqlglot-backed SQL validator judged by Sonnet 4.6, executor, and interpreter. Each stage is inspectable.
Plug in a database, a warehouse, or a SaaS source via OAuth or credentials. Read-only by default.
We profile your data and draft a semantic layer — table and column descriptions, joins, example questions. You edit it once; every answer after gets sharper.
Type a question the way you'd ask a teammate. A schema selector narrows the surface, a planner drafts the query, a validator proves it's safe to run.
The executor runs the plan against your source and an interpreter picks the right visualization — chart, KPI, or table. The full plan ships with every result.
Built by Algoscale
AnalystIQ is built by Algoscale, a data engineering and AI services firm. We've spent the last decade building pipelines, warehouses, and analytics stacks for teams who needed answers faster than their tools could deliver. AnalystIQ is our first SaaS — the product we kept wishing we could hand our clients at the end of every engagement.
Principles
Three commitments that shape every product decision we make.
Every connector defaults to scope=nothing. You opt in to specific tables and fields — anything outside that boundary is invisible to the LLM, the planner, and the executor.
We don't store result rows. Query plans are persisted for auditability; the rows themselves live in a short-TTL cache and never leave your region.
Every answer is accompanied by the exact SQL plan we ran. Analysts can verify the logic, copy it into their own tools, or hand it to a teammate for review.