Novel
FRESH DISCOVERYAI that builds on its own discoveries, round after round. First it spots patterns, then digs into why they matter, then connects findings into recommendations.
- Aware of your KPI targets
- Auto chart selection
- Round-by-round depth
Don't wait on analysts ever again.
Upload a CSV. Get the chart, the number, and the sentence you were going to ask someone for. Three modes — Novel, Blueprint, Meta — covering one-off questions, recurring drops, and cross-file synthesis.
00 sec
Setup time
12 sec
Median analysis
100k+
Files processed / mo
3 modes
Novel · Blueprint · Meta
One file in. The right kind of analysis out.
Blueprint
REPLAYTurn your best analysis into a one-click recipe. Same lens, new data, comparable insights every time. Monthly and quarterly reporting becomes effortless.
- Reusable recipe
- Lives inside Silos
- Team standard
Meta
SYNTHESISYour past analyses, finally in conversation with each other. Synthesize across many saved analyses — cross-domain or time-lapse — into one boardroom-ready dashboard.
- Complementary mode
- Additive (timeline) mode
- No re-uploading
How a file becomes a finding.
One ingest. Three lanes. Same output shape — chart, finding, report. Switch lanes any time without reuploading.
CSV / XLSX
↓
INGEST
↓
INGEST
NOVELfresh discovery →
BLUEPRINTreplay →
METAsynthesis →
CHART
FINDING
REPORT
Trust the data before you trust the finding.
The instant your file lands, Analitika standardizes column names, validates the file format, profiles every column, and surfaces a transparent diagnostics panel. Rule-based, deterministic, and non-destructive — the AI isn't in the cleanup loop.
DATA REVIEW · Q3-REVENUE.CSVPARSED · 0.4s
Rows
4,128
Columns
12
Quality
Excellent
AUTO-FIXED
What we quietly repair
- Column names — spaces → underscores, $ → usd, % → pct
- Special-character stripping, numeric-prefix repair
- File-format validation via magic bytes
- Tier-aware row truncation
SURFACED
What we honestly diagnose
- Missing values with column-level heatmap
- Single-value and high-cardinality columns
- Type mismatches and temporal viability
- Hidden identifiers and out-of-pattern entries
Rule-based · deterministic · non-destructive — the AI never touches the cleanup loop
Set it once. Recognized forever.
Every file you upload gets fingerprinted into a Data Label — a memorable identity like "Monthly Sales Report" or "Customer Survey". The first time, the app learns it. Every time after, the app recognizes it — and the right template, the right targets, and the right context snap into place automatically.
Output Template
TEMPLATEYour preferred report layout, brand kit, and chart types — already applied when the file lands.
KPI Targets
TARGETSPerformance thresholds the AI is aware of from round one. Targets travel with the data, not the file.
Analysis Context
CONTEXTHow you like the analysis presented — angle, depth, audience. The AI starts where you left off.
01
First upload → Label is created. You set template, targets, and angle once.
02
Second upload → Label is recognized. Everything snaps in automatically.
03
Hundredth upload → Same one click. Same recognized label. Zero re-setup.
Four steps. None of them involve emailing anyone.
01
Upload
Drag and drop your CSV or Excel file. No connectors to configure.
02
Configure
Pick an analysis angle, add business context, map KPIs to your targets, choose how deep to go.
03
Watch it work
A progress indicator shows the AI advancing round by round, building on each discovery.
04
Review
Interactive charts, written insights, downloadable HTML report. Share the link.
An example round, sketched at 1:1.
The output is intentionally legible: chart, headline finding, three supporting metrics. Export as HTML and forward.
Round 03 Analysis
Round 3 Insights
Key Findings
Q3 revenue is up 18.4% YoY, with the strongest lift concentrated in the final two months of the quarter. September alone cleared the 35,000 target, and the segment mix shifted toward higher-value deals — the trend line points to continued momentum heading into Q4.
📈 Round 3: Primary Analysis (Charts 1–2)
Revenue by month
Revenue
Revenue by segment
$4,820
Avg Deal
3.81%
Conv. Rate
$214
CAC
Pro Tier Analysis · Completed 3 rounds with 6 total charts
Three tiers. Same features. Pay for usage, not lock-outs.
FULL PRICING →| Starter$25 | ● MOST CHOSEN Standard$30 |
Pro$48 | |
|---|---|---|---|
| Per seat / month | $25 | $30 | $48 |
| Novel rounds/mo | 20 | 40 | 80 |
| Blueprint rounds/mo | 60 | 120 | 200 |
| Meta-analyses/mo | 30 | 45 | 60 |
| Tags·Targets·Templates·Silos | Unlimited | Unlimited | Unlimited |
| SSO·Silo sharing·PDF export | ✓ | ✓ | ✓ |
| Choose Starter → | Choose Standard → | Choose Pro → |
Questions, answered.
Where does my data go?
Files are processed in-memory and dropped after analysis. Findings and blueprints are stored in your account; raw data isn't. VPC-isolated processing is available on request.
Which model powers the analysis?
Anthropic's Claude. We pin a model version per release so blueprints replay deterministically. The model never sees your file headers or rows outside an active analysis run.
Can I change the chart the model picks?
The AI selects the chart shape it thinks best fits each finding — you can't edit the chart directly today, but you can redirect the round to ask the question a different way and get a different visualization. Chart editing is on the roadmap.
What does 'rounds' mean?
Each round is a deeper pass — the AI revisits your data armed with everything it just discovered. More rounds means a richer story. Your tier determines how many Novel and Blueprint rounds you can run per month, not what's locked behind a paywall.
How is Meta different from Blueprint?
Blueprint replays one analysis on a new file. Meta synthesizes findings across many already-run analyses — useful for a year of weekly drops, or ten regions running the same blueprint.
Do I need to install anything?
No. Analitika runs entirely in the browser — drop a file in, read the finding out.
Can I export the report?
Yes — HTML, PDF, or individual PNGs. The HTML report is self-contained; you can email it or commit it to a repo.
What file types are supported?
CSV and Excel files on every tier — no feature lock-outs by plan. The only thing tiers change is how many analyses you can run per month.
What about Tags, Silos, Targets, Templates?
Unlimited on every tier. Silo sharing, SSO, and PDF export are also included for everyone — pricing is purely about usage volume.
Join our waitlist.
The app isn't open to the public yet. Add your email to the waitlist and we'll let you know the moment you can upload your first file.
Join the waitlist →