Product analytics on the data you already own — with an AI analyst that investigates the why.
Ask it why, not just what
Plain-language questions answered in seconds — the drop found, the segment named, the release blamed, and the chart it all came from attached.
See the AI analyst 02The tool your PMs already know
Funnels, retention, metrics, journeys, dynamic segments and alerts — the full product-analytics seat, fluent on day one, running on your data.
See the product 03Privacy-first architecture, at 50–70% less
Deployed inside your account, reading your warehouse in place. No second copy of your events — which is exactly why the bill falls.
See how we deliverEvery product team has more questions than analyst hours. This answers them — and shows its working, every time.
No SQL, no ticket, no waiting for someone to have a free afternoon. The analyst finds the drop, isolates the segment, names the likely cause — and shows you the chart it read.
Most questions a product team asks are simple — and still wait a week for an analyst. The AI analyst answers those in seconds, with receipts. Your analysts get their time back for the work that deserves it.
Shipping in this order: catalog → Ask → MCP → Watch → Dig. We'll tell you on the call exactly what's live the week you'd go in.
Point a language model at raw tables and it guesses: at the column that sounds right, at what “active” means, at a number nobody can check. Ours is never asked to. Every answer resolves through one catalog of your metrics, dimensions, cohorts and context — drafted from your own warehouse, then signed off by someone on your side.
Schemas, tables, columns, volumes, relationships and how they are actually populated — profiled where they already sit, read-only. It learns that plan_tier holds four real values, not that it sounds like a plan.
From that index plus your own docs, it proposes the metrics your team argues about, the dimensions each can legitimately be split by, the cohorts that matter, and the business context behind them. Every definition is run against real data before it is offered to you.
Your analyst edits the wording, fixes the edge case, rejects what is wrong. Nothing enters the catalog without that sign-off — and nothing outside the catalog ever becomes a number. Definitions stay readable text you can diff and revert.
An AI analyst is only as good as the tool underneath it. This is the full product-analytics seat — the one your team already knows how to drive.
Trends, funnels, retention and journeys on one board, reading your events where they land. Nothing was re-instrumented and nothing was copied out to build this.
A replacement for the product-analytics seat: funnels, retention, metrics, journeys, dynamic segments, alerts. Fluent on day one for a team that knows Mixpanel or Amplitude. Not session replay, not feature flags.
Privacy-first by construction, not by policy. The architecture that keeps your events in your cloud is the same one that takes 50–70% off the bill.
A copy of every event leaves your cloud. You pay to store it twice — so you ration what you track.
The analytics moves to the data. No event data crosses the boundary.
If your events don't land in a warehouse yet, there is a collection path — we'll tell you which one you're on before the call ends. · Prefer not to operate anything? There is a hosted option, with a different trade-off on where the data sits — deployment options →
No compliance badges until we hold them. DPDP-ready because of the shape of the system; SOC 2 status on the Security page. Full Security page →
Design-partner pricing on the call. [confirm: unit — per workspace · per seat · flat tier]
50–70% is against incumbent list pricing at your event volume; the worked example with assumptions is on the call.
| Mixpanel / Amplitude | Self-hosted OSS (e.g. PostHog) | Build it in-house | DepSmith | |
|---|---|---|---|---|
| Where your event data lives | Their servers | Yours, if you operate it | Yours | Yours |
| Infrastructure you operate | None | The full stack — ClickHouse, Kafka, the rest | All of it | None — deployed into your cloud |
| What you are billed for | Data retention | Infrastructure and the team to run it | A platform team | The software |
| Data residency under your control | ✗ | ✓ | ✓ | ✓ |
| AI analyst | Assistant over their copy of your data | ✗ | Build it | Verified catalog, in your cloud — app · Slack · MCP |
On building it yourself. It works — teams have done it. The price is a dedicated platform team, the better part of a year, and a standing capability gap versus the tool their PMs had before. That team is the cost.
We will tell you so on the call.
Read-only, to where your events already land. Inside your cloud.
The analyst reads your events and your docs and drafts your metrics, cohorts and funnels — each verified against real data.
A person on your side signs off each definition. Nothing becomes a number without this step.
Dashboards on day one; questions in the product, Slack or MCP as each ships.
Typical: days, not a quarter, when your events already land in a warehouse you control. Your history is queryable from the first day — nothing is backfilled.
No. DepSmith is deployed inside your cloud account and reads your events where they already land. We never take a copy and never hold your data. What the AI layer sees, and which model runs it, is on the Security page. [confirm: provider, BYO-key option, what leaves the VPC]
Yes, with one more step. If your events don't already land somewhere you control, there is a collection path that gets them there first. We'll tell you which path you're on before the first call ends.
Yes, when your events already land in a warehouse you control — we read what you hold, so there is no backfill and no gap. If your history lives only inside Mixpanel or Amplitude, there is an import step; we'll confirm the path for your setup on the call.
For the product analytics your PMs use daily — funnels, retention, trends, journeys, cohorts, dashboards — yes. It is built to be fluent on day one for a team that knows those tools. It is not session replay or feature flags.
The AI analyst proposes it from your data and your docs — metrics, dimensions, cohorts, funnels — and verifies each against real data. A person on your side approves before anything is trusted. Nothing becomes a number without that step.
Self-hosting solves residency and hands you a stack to operate — ClickHouse, Kafka, the rest — plus their SDK to adopt. Here you operate nothing and the data doesn't move. The comparison table above has it row by row.
Days, for the common case where your events already land in a warehouse. The setup path with time-to-live per step is on the How it works page. [confirm: steps and durations]
Nothing to export. The data was always yours, in your cloud; the catalog definitions are readable text. You switch off the deployment.
Bring your warehouse. In thirty minutes we'll answer one real question from your data, look at where your events live today, and tell you honestly whether this is a fit yet.
Not ready to talk? Leave an email and we will send you the occasional real update — no drip sequence.