WEEK 1 · ROBO-BEAR
FACTORY MAX: 100 TOYS
☀️ Week 1: Sunny Opening Day!
THIS WEEK'S KIDS FORECAST68 toys
WAREHOUSE + HIGHWAY TRUCKS117 toys
NEXT WEEK'S FORECAST76 toys
YOUR KNOWLEDGE HUB
What would you like to understand?
Take a course, explore a planning idea, or put an assumption to work.
LEARN
The forecasting course
Five connected tracks, 50 lessons and a clear learning path.
Start learning →PRACTISETools & games
Explore forecasts, scenarios, inventory and planning decisions.
Choose a tool →THINK FURTHERArticles & ideas
Browse practical briefs and longer field guides.
Find an idea →Search all lessons, guides, tools and AllIdeas →
—COURSE OVERVIEW · 499-HOUR LEARNING PATH
Supply Chain Forecasting Course
Learn how a demand signal becomes a supply plan, how real limits change that plan, and how teams turn it into a production schedule.
THE PLANNING PATH01 → 05
Your course progress: 0 of 5 lessonsMark lessons complete as you go. Progress stays in this browser.Learning points: 0 · Badges: earn one by learning · No cash value
END-TO-END CAPSTONE · PUT THE PLAN TOGETHER
A shortage before the Friday shipment
Use one realistic scenario to connect the forecast, inventory, supplier limit, line capacity and customer promise. Pick a response to reveal the trade-offs.
DEMAND TO COVER120 forecast + 8 backlog128 bottles in total
AVAILABLE SUPPLY28 stock + 20 receipts48 already cover demand
NET BUILD NEED80 bottles128 − 48 = 80
Main supplier72 kits
→Normal line76 units
→Dispatch cap82 units
Supplier is the first constraintWhat would you do?
Assumptions: alternate kits cost ₹35 each above base, overtime costs ₹90 per extra unit, and dispatch can handle 82 units. These are made-up teaching figures, not market prices. Whether the premium is worthwhile depends on the cost of a missed order and available policy.
Learning exercise: the quantities are illustrative. A real plan also depends on product rules, calendars, data quality, supplier agreements and local operating constraints.
ANIMATED SUPPLY-CHAIN ADVENTURE · AGES 10 TO 100
Supply Plan Sprint: Toy Factory Express
Watch your Supplier, Toy Factory, Smart Warehouse, and Delivery Trucks come alive! Balance happy kids, safety stock, and coins across 5 action-packed weeks.
60FPS ANIMATED · 5 WEEKS
Kid-Safe Educational Game · 100% private in-browser play with fictional toys and coins. No personal data or email required. Click any building or truck on the animated map to inspect it!
INSTANT PLAY ACCESS
Play the animated adventure or sign in for ranked play.
Play the 5-week animated Toy Factory Express game right away—or sign in if account services are enabled.
Interactive 60FPS HTML5 Canvas Supply Chain World
Kid-safe privacy: your call-sign stays on this screen. Never enter personal information.
ADVENTURE COMPLETE!
SUPPLIER 🌲 → FACTORY 🏭 → WAREHOUSE 📦 → TOY SHOP 🏪
Ready to run the Toy Factory Express?
Every week, check the weather forecast and choose how many toys to build. Keep extra Safety Stock in your warehouse so surprise birthday parties or rainstorms never leave kids waiting!
WEEK 1 / 5
SCORE 100 ⭐
HAPPY KIDS 100% 😊
YOUR WEEKLY PLAN
Trucks take 1 week to arrive
1. Choose how many toys to build for next week
Your Captain Score: 100 ⭐
SUPPLY CHAIN IDEAS · KID-SAFE EDUCATIONAL EDITION
100% ORIGINAL VECTOR CANVAS & SYNTH AUDIO · ZERO TRACKERS
QUANTITATIVE FINANCE + SUPPLY-CHAIN SCIENCE
SCIENTIFIC ENGINE v2.0
FBS — Scientific Financial & Inventory Forecaster
An econometric and working-capital decision laboratory that couples a supply-chain inventory forecast with a 5-year historical balance sheet using State-Space Damped Holt, Theil-Sen robust regression, and Monte Carlo P10–P90 uncertainty bands.
100% Private In-Browser Computation. PDFs are parsed locally on your device and are never uploaded to any server. Or test immediately with a 1-Click Verified Benchmark Dataset below. Educational & analytical decision support only—not statutory accounting, tax, or investment advice.
COUPLED WORKING-CAPITAL & STATE-SPACE MODELING
Why inventory plans and financial statements must be solved together.
Conventional forecasts project inventory in isolation from trade payables, working-capital cash absorption, and holding-cost drag. FBS v2.0 reframes the problem scientifically: it backtests three time-series estimators out-of-sample on your 5-year history, couples next-year inventory shifts to supplier payables (DPO/DIO elasticity) and carrying costs (WACC + storage + obsolescence), and quantifies P10–P90 confidence bands.
Blends Damped Holt State-Space trend, Theil-Sen non-parametric slope, and historical drift weighted by walk-forward out-of-sample RMSE.
Quantifies how inventory buffers or lean releases alter Net Working Capital (ΔNWC), supplier payables financing, and retained earnings.
Outputs 80% confidence intervals, Current/Quick/Debt-to-Equity ratios, and a 6-year visual trajectory fan chart.
01Load Data or 6 PDFsUpload 5 balance sheets + 1 inventory PDF, or load a 1-click benchmark
→
02Verify & Tune LeversAudit the 5-year accounting identity and set WACC / DPO coupling
→
03Analyze Ensemble & Fan ChartInspect P10–P90 projections, backtest WAPE, and NWC cash impact
STEP 1 · LOCAL PDF EXTRACTION OR INSTANT BENCHMARK DATASET
Choose source reports or load a verified benchmark company
Upload five annual balance-sheet PDFs (oldest to newest) plus one inventory forecast PDF—or click a benchmark company below to test the scientific forecasting suite immediately.
Upload five annual statements, one PDF for each year.
pdf only · five files · maximum 10 MB eachChoose a report with a projected inventory line.
pdf only · one file · maximum 10 MB
Use the same unit for all uploaded reports and values.
Select all six PDFs—or load a 1-click Scientist Benchmark Dataset above.
What happens to my PDFs?
The PDFs and extracted text stay in your browser session. The tool does not send them to a server or save them to an account. Closing or refreshing the tab clears the selected files and extracted values. Avoid uploading confidential records to any website unless you have permission.
Check extracted figures and tune working-capital coupling
Verify every historical line item below. Then adjust the Scientific Model parameters to test how inventory buffers, holding costs, and supplier terms impact the projected balance sheet.
Document order: oldest year in the first column, newest year in the fifth. Blank values are treated as missing, not zero.
Enter or verify the historic values to check each year’s balance.
SUPPLY-CHAIN INVENTORY PLAN INPUT
Next-year baseline inventory from your report
Review the relevant forecast period in the source PDF, then enter its baseline inventory figure here.
Ensemble weights models by inverse walk-forward RMSE on Years 3→4 and 4→5.
Simulate lean inventory release (−30%) vs. safety buffer build (+30%).
Models P&L / retained-earnings drag from incremental inventory holding.
Share of inventory change financed organically via trade payables.
Inspect locally extracted PDF text / dataset notes
Projected Balance Sheet, P10–P90 Bands & Liquidity Scorecard
| Balance-sheet line | Projected (P50) & P10–P90 Band | Scientific Estimation Method |
|---|
Scientific methodology, equations, and governance limits
COUPLED OPERATIONS & FINANCE ARCHITECTURE
From supply-chain inventory posture to balance-sheet resilience.
Rather than treating balance-sheet lines as isolated random walks, FBS v2.0 links operational inventory decisions directly to working-capital financing (Trade Payables DPO/DIO elasticity), holding-cost drag (WACC + warehousing + obsolescence), and out-of-sample backtested time-series ensembles.
01 / PRACTICE LIBRARY
100 ARTICLESSupply Chain Articles
100 original short reads on practical supply chain ideas, AI and operations.
IN-DEPTH FIELD GUIDES · SOURCE-LINKED
Browse all field guides →Go beyond the quick read.
Longer practical guides expand ideas from the library and separate source findings from practitioner commentary.
FORECASTING · 8 MINForecast accuracy without score-chasingChoose metrics, spot bias and validate with time-ordered holdouts.INVENTORY · 9 MINSafety stock and reorder pointsUnderstand the assumptions behind a buffer and when it needs review.INTEGRATED PLANNING · 10 MINFrom forecast to a feasible production planConnect uncertain demand to constrained master scheduling.OPERATIONS · 8 MINFinite capacity and practical sequencingUse real available time, bottlenecks, due dates and setup trade-offs.
04 / TECHNICAL LIBRARY
50 ARTICLESSupply Chain Forecasting · Advanced
50 technical reads on probabilistic forecasting, machine learning and forecast value.
SEARCH RESULTS
Explore the library
Results across all collections.
EDUCATIONAL SIMULATOR · CONTROL TOWER
Six weeks. One living supply chain.
Make a plan with incomplete information, then see how lead times, capacity, variability and disruptions change the result.
Learning simulation · All costs, demand and events are illustrative. This is not a forecast or operational planning tool.
Play without an account, or sign in to save your decisions and resume later.
WEEKLY DECISION
WEEK 01Week 1 · Check the signal
DISRUPTION / MARKET SIGNAL
Demand outlook
Event notes appear here.
FORECAST100units
SUPPLIER RECEIPTS DUE0 units
FACTORY CAPACITY THIS WEEK100 units
PIPELINE0 units
Arrives next week · $4/unit · capacity constrained
0 units100 units
Available now · $11/unit · limited emergency capacity
0 units40 units
💡
Planner's note: Base production is cheaper but has a one-week lead time. Expedite can protect service today, but costs more and has a 40-unit cap.
Material flow this week
UNIT BALANCEOpening stock90+Supplier receipts0+Expedite0−Demand served0=Closing stock90
Backorders are tracked separately and carried into the next round.
FORECASTING PLAYGROUND
Change the demand shape.
Adjust trend, seasonal swing and noise. See how a simple baseline compares with a small AI-style model on held-out weeks.
Read the calculator guides →Illustrative only · The “AI-style” forecast is a local regression model with a trend and seasonal terms. It is not an LLM, does not upload data, and is not a production forecast.
Demand scenario
32 simulated periods · 24 train · 8 holdout · no external data
Forecast vs. observed
HOLDOUT BACKTESTActualSimple moving averageAI-style trend + season model
PLANNING CALCULATORS
ILLUSTRATIVE · NO DATA UPLOADTurn the signal into a simple plan.
Try the numbers in your browser. These tools use transparent starter formulas, not a full planning system.
Score a forecast
Compare actual and forecast demand, in matching time periods and units.
MAE is in units; WAPE shows total absolute error relative to total actual demand; signed bias is forecast minus actual. WAPE is undefined when total actual demand is zero.Reorder point & suggested order
A basic continuous-review example: expected lead-time demand plus a chosen buffer.
Reorder point = average demand × lead time + buffer. Suggested order restores inventory position to that point; lot sizes and calendars are not modeled.Check a line load
Estimate required run and setup minutes against the time actually available.
Includes one setup allowance and one item rate. Real schedules may need multiple routings, shifts, labor rules, downtime and sequence-dependent setup.Educational only. Validate formulas, units, policy settings and service targets with your planning team before using them for operational decisions.
Good yardstick, slow to turn.
A moving average is transparent and hard to overfit. It can lag after a trend shift, flatten peaks, and miss recurring seasonal peaks.
Can fit a pattern that vanished.
A trend plus seasonal model can be wrong after a regime change, with short or biased history, or when promotions and supply constraints reshape observed sales.
Model names are descriptive of this demonstration. Compare on rolling, time-ordered holdouts before selecting any method for real planning.
Contact Founder
LinkedInYOUR PRIVATE SIMULATOR SAVEPick up where
Pick up where
you left off.
Create an account or sign in with your email to keep your six-week simulation progress. Your saved game can be opened on another device using the same account.
01
Save weekly decisions and results automatically.
02
Resume your in-progress scenario after signing in.
03
Recover access through an email password reset link.
Only simulator progress is stored in your account. Reading-list preferences remain in this browser.
EMAIL ACCOUNT
Welcome back
Sign in to sync your simulator progress.
Preparing secure account service…
Signed in as
Your simulator saves automatically after each decision. Sign in with this email on another device to resume.
Email accounts are protected by Supabase authentication. The service must be configured by the site owner before account features are available.
PRIVACY & EDUCATIONAL USE
Clear about data. Clear about limits.
Last updated 29 October 2026.
Contact
The Contact CEO page links to Akshay Kaushal’s public LinkedIn profile. This static website does not collect messages or contact form submissions.
Account and simulator data
If account features are configured, Supabase handles passwords, email verification, and password resets; this static website does not store plaintext passwords. Saved simulator progress is protected by row-level security so a signed-in user can access only their own saved record. The site does not ask for sensitive business data. Other site features may continue to use email sign-in.
Local browser data
The reading list and simulator best score remain in your current browser using local storage. Clearing browser storage removes them. If cloud saving is unavailable, the screen displays a save error; do not assume the latest changes were retained. FBS PDF files and extracted figures are processed in the browser and are not uploaded or retained by this static site.
Third-party services
Bootstrap, the PDF.js reader and the Supabase client are bundled locally at pinned versions. PDF JavaScript and XFA are disabled while reading files. System fonts are used so text renders without a third-party font request. No analytics or advertising trackers are included in this static package. When enabled, authentication and simulator saves use the site owner’s Supabase project. External research and LinkedIn links open pages operated by those providers, whose own privacy terms apply.
Educational use
The simulator, forecasting playground and articles are educational. Their scenarios and results are illustrative and are not operational, financial or professional advice. Validate methods against your data, constraints and approved planning process.