What are the opening hours?

Monday 9am - 5pm Tuesday 9am - 5pm Wednesday 9am - 5pm Thursday 9am - 5pm Friday 9am - 5pm

dotData

Provide end-to-end data science solutions to help enterprise leverage data for business innovation.

dotData helps businesses that are just getting started with predictive analytics and more mature organizations that have established data science and data engineering processes. Our core technology automatically converts data from data warehouses and lakes into data marts and feature tables by exploring the relationships between varied data tables with hundreds of columns and millions of rows.

10/22/2026

We’re excited to share that dotData was recently featured in AutoFin Journal, offering strategic guidance on how auto lenders can safely deploy AI without exposing their portfolios to hidden credit risks or regulatory penalties.

Featured Articles

1. The LLM Compliance Paradox & Separating Conversational AI from Quantitative Underwriting
By Joshua Gordon, Ph.D., Senior Data Scientist at dotData

Read the article➡️ https://hubs.ly/Q04BR6FS0

2. Calibrating AI Risk Models for Shifting Macroeconomic Cycles
By Bart Blackburn, Ph.D., Staff Data Scientist at dotData

Read the article➡️
https://hubs.ly/Q04BR6D00

10/20/2026

A mispriced contract within a promotion still books and still looks profitable because the manufacturer covers the gap. The pricing model never gets tested there.

The test that exposes it: back-test the promotional contracts as if no subvention had been paid, and list the tiers that would have come out underpriced. Those same tiers are pricing today's standard-rate book, where the annualized flow of auto balances into serious delinquency reached 3.00% in Q2 2026, according to the New York Fed.

Read the full analysis ⤵️

10/15/2026

Blended averages are the enemy of an effective risk strategy.

In the current auto lending market, we’re seeing prime 60+ DPD rates at 0.42% while subprime has hit 6.90%. That is a signal 15x higher than prim, a distinction that is almost entirely lost if your vendor demonstration relies on portfolio-wide averages.

If your decisioning software cannot handle raw relational data to construct payment-to-cash-flow relationships, it’s not just inefficient, it’s missing the most critical risks in your portfolio.

Don't let marketing claims mask the performance gaps in your lending architecture. We’ve identified where standard evaluations fall short and how to demand better data handling during your next vendor demo.

👉 Check the guide now: https://hubs.ly/Q04BdFrq0

10/13/2026

Yokohama Rubber and dotData Begin Joint Development of an AI Agent for Manufacturing R&D

Read the announcement 👇
https://hubs.ly/Q04zYkfs0

10/08/2026

When near-prime volume slides and charge-offs stay flat, look at the buy rate before dealer compensation. A borrower quoted too high a rate never defaults; the dealer sends the contract to a lender with a better number, and it never books.

According to a paper cited by the Congressional Research Service, 83% of auto loans are obtained indirectly, so the leak can reach most of a large indirect book. Pair look-to-book with losses by tier; a group with low losses and shrinking volume is probably overpriced.

Read the full analysis ➡️ https://hubs.ly/Q04zp5W10

10/06/2026

In 2026, "AI in Lending" is a bloated category covering 3 completely different jobs. If you evaluate them under a single RFP scorecard, your project will quietly fail.

1️⃣ Document Automation → Solves intake speed (Validates documents, not credit risk)
2️⃣ Decisioning Automation → Solves queue throughput (Multiplies model blind spots)
3️⃣ Signal Discovery → Solves portfolio drift (Uncovers non-obvious KPI drivers)

Stop buying a document scanner when you need a risk discovery tool.

Learn more 👇

10/01/2026

Tier-level pricing treats every borrower in a band as the same borrower.

That assumption holds near the center of a tier. At the edges, expected loss diverges sharply from actual loss and no credit score captures that gap automatically.

In Q2 2026, the average rate was 6.35% for new vehicles and 11.19% for used ones. The spread widens further below prime, where edge-of-tier mispricing does the most damage.

The measurement gap is the real problem. Rank borrower groups by the difference between expected and actual loss. The widest gaps are where your rate sheet is wrong in both directions simultaneously.

dotData surfaces the Driver Signals that explain why actual loss diverges from forecast, without requiring a model rebuild or a new data pipeline.

Learn more ⤵️

09/29/2026

The NCO rate is the P&L's final verdict on underwriting decisions made 18 to 36 months earlier.

By the time it moves, the capital has already left the building. Auto lenders are recovering only about 30 cents for every dollar through repossessions — meaning Loss Given Default is materially more severe than headline charge-off rates suggest.

Over 53% of used-vehicle loans were issued with LTV ratios above 120%. The collateral position does not absorb the loss. The P&L does.

The Federal Reserve reported consumer loan charge-off rates hit 2.87% across all commercial banks in Q4 2025. On a $500 million portfolio, that is not a rounding error.

dotData Signal Intelligence identifies which specific accounts are trending toward default within 6 to 12 months — before the charge-off report confirms what mitigation can no longer address. The output is SQL-ready. Deployment isn't measured in quarters.

The window to act on 2023 and 2024 vintage deterioration is narrow.

Request a portfolio review: https://hubs.ly/Q04ywY6P0

09/23/2026

Four-tenths of one percent of a portfolio defaulted at 50.0%. The portfolio average never moved.

Most risk models are calibrated to read the book at the aggregate level. That works until the deterioration stops being evenly distributed — and it rarely is. Loss concentrates. It concentrates in slices too small to shift a mean and too specific for a standing scorecard to have tested.

Here is what that looks like in a real auto portfolio. A secured credit card Driver Signal covers 4.93% of records and lifts the 90-day-past-due rate from a 20% baseline to 39.3%. An education loan Driver Signal covers 5.84% and lifts it to 25.6%. Both are moderate. Stack them, and the 0.443% of the portfolio meeting both criteria carries a 50.0% default rate.

A 30-percentage-point surge, invisible to the average.

A quant can find this by hand. It takes one hypothesis at a time, and there are millions of combinations. Signal Intelligence runs the search continuously on loans already booked, upstream of whatever decisioning stack the lender operates.

See which segment of your booked portfolio is drifting:
👉 https://hubs.ly/Q04xWS0Q0

Read more below ⤵️

09/17/2026

Stop letting flawed software evaluations dictate your risk architecture

Most auto lending risk leaders are set up to fail before a vendor demo even starts. Why? Because traditional evaluation matrices lump four fundamentally different categories into one generic label:

🔹 Origination Fraud Filters
🔹 Decisioning Engines
🔹 BI Dashboards
🔹 Signal Intelligence Platforms

When you score completely different software using the same criteria, your "winner" is selected by pure accident, leaving critical gaps in your lending ecosystem.

Before you launch your next procurement cycle, learn how to properly segment and evaluate your software stack to protect your portfolio.

👉 Fix your risk evaluation gap here: https://hubs.ly/Q04xJBxn0

Address

1730 S. El Camino Real, No 200
San Mateo, CA
94403

Opening Hours

Monday	9am - 5pm
Tuesday	9am - 5pm
Wednesday	9am - 5pm
Thursday	9am - 5pm
Friday	9am - 5pm

Website

https://dotData.com/

Alerts

Be the first to know and let us send you an email when dotData posts news and promotions. Your email address will not be used for any other purpose, and you can unsubscribe at any time.

Contact The Business

Send a message to dotData:

Want your business to be the top-listed computer & electronics service in San Mateo?