BlindSpot

Find your blind spots before they cost you.

You’re already doing the hard part — tracking the data, making the calls, running the numbers. BlindSpot structures what you’ve built, finds the patterns you can’t see from inside, and shows you exactly where your edge lives and where it doesn’t.

Your Data→Fact Tables→Dimensions→Dashboard→Blind Spots

↓ the data changes · the engine doesn’t care ↓

4Domains

1Architecture

0Lines of Production Code

$0API Costs

Domains

Same engine. Different data.

Each BlindSpot module uses identical architecture — star schema, fact tables, dimensional slicing.

Trading & Betting

BlindSpot Trading

Strategy backtesting, bet tracking, and performance analytics. Built for a trader who returned 117% from a notebook — now structured so you can see where the edge actually lives.

Strategies

A / B

Bet Types

ATS · ML · O/U

Economy

10K units

backtestingMarch Madnessstrategy comparisonbankroll tracking

Explore →→

⚔ D&D Campaign Intelligence

BlindSpot Campaign

Your campaign is generating data — rolls, quest threads, spotlight balance, faction heat. You're just not reading it. BlindSpot tracks everything and shows you what you're missing before session 6 kills it.

Tracked

347 rolls

Quests

12 active

Briefs

LLM-gen

dice analyticsspotlight balancequest velocityfaction heat

Explore →→

🎮 Steam Library Analytics

BlindSpot Steam

247 games. $3,847 spent. 42% unplayed. Your Steam library is a dataset — with genre bias, impulse patterns, and a backlog that grows faster than you play. See where your money actually goes.

Library

247 games

Blended $/hr

$1.35

Unplayed

42%

$/hour efficiencygenre analysispurchase decisionsbacklog priority

Explore →→

⚡ Local LLM Integration

BlindSpot × LLM

The engine builds the data. The model reads it. Local GPU inference via Ollama — intelligence briefs, transcript parsing, convergent analysis from multiple models. Zero cloud. Zero API costs. Your hardware.

Runtime

Ollama

Models

3 loaded

Cost

$0 / ∞

GPU inferenceconvergent analysisgoverned promptsprivacy-first

Explore →→

Architecture

One engine. Every domain.

Star schema → fact tables → dimension tables → dashboard → blind spots

Data Sources

trade logs

session transcripts

Steam API

CSV imports

manual entry

Fact Tables

Fact_Trades

Fact_Rolls

Fact_Quests

Fact_Games

Fact_Purchases

Dimensions

Dim_Strategy

Dim_Ticker

Dim_PC

Dim_Faction

Dim_Genre

Dim_Game

Dim_Date

Intelligence

LLM parsing

convergent analysis

intelligence briefs

governed prompts

Output

dashboards

blind spot flags

decision engines

pre-session briefs

The data changes. The engine doesn’t care.
Same star schema. Same dimensional slicing. Same pipeline. Different domain.

Questions · BlindSpot Answers

across all domains

Trading

"Am I better against the spread or on moneylines?"

ATS: 62% · ML: 48%

Trading

"Which strategy is actually making me money?"

A: +$2,890 · B: +$1,390

Campaign

"Is the spotlight balanced or is someone fading?"

62/38 split · flag raised

Campaign

"Are the dice actually cursed?"

avg 8.7 · p=0.04 · yes

Steam

"Should I buy Civilization VII?"

82% chance unplayed · HOLD

Steam

"Where is my money actually going?"

RPGs: $682 · 19% played

LLM

"Do both models agree on this flag?"

converge: yes · high confidence

Any

"What am I missing?"

that's the whole point

The notebook can’t slice.

You’ve been tracking things — trades in a spreadsheet, sessions in your head, purchases on impulse. You’re already doing the work. But the format you’re using can’t show you what you’re missing.

BlindSpot takes the data you already have and structures it so you can cut it by dimension — by strategy, by genre, by PC, by time period. Same architecture every time. Fact tables. Dimension tables. Dashboards. Flags.

The engine was built for enterprise audit coverage analysis. Turns out it works for anything with structured data and decisions that have consequences. The data changes. The engine doesn’t care.

BlindSpot doesn’t tell you what to do. It shows you what you can’t see from inside. The rest is your call.

— Dave Kitchens

Dropdown Logistics · same engine · different data

BlindSpot · find your blind spots before they cost you

built with the same engine that doesn’t care where the data comes from

Trading Campaign Steam LLM