01 · About
Why bring me your next decision?
Open
02 · Leadership
Teams pointed at one decision.
Open
03 · Hands-on
Still hands on, end to end.
Open
04 · Research
Why good models go unused.
Open
05 · Background
Where the judgment comes from.
Open
06 · Services
Bring the decision, not just the data.
OpenRichard Henderson
AI leader — LEADS AI TEAMS
- AI leader — LEADS AI TEAMS
- Data scientist — ML · LLMS · RECSYS
- Decision science researcher — WHY MODELS GO UNUSED
02 · Leadership
One team. One decision.
I bring every discipline into focus on the decision that matters.
Read the playbook- Engineering
- Analysts
- ML
- Product
03 · Hands-on
I still build it myself.
From raw data to something a person can act on.
Fly into the Lab- Projects 15
- AUC 0.848
- Uplift +$636K
04 · Research
Why good models go unused.
Accuracy isn’t adoption. I study what sits in between.
05 · Background
Where the judgment comes from.
Four years in the U.S. Navy, then pricing, data science and now doctoral research.
- 2Master’sUNT
- 1FinTech cert.SMU
06 · Services
Bring me the decision.
Press a way to work together. The print shows where I come in.
Stages3/4
- 01 · Idea accuracy ≠ adoption
- 02 · Framing decision first
- 03 · Adoption use, not logins
- 04 · Value test could have said no
The Lab
The Lab: 18 live demos
01 · Models
Lending decision lab
50,000
synthetic members
02 · Models
Subscriber value lab
+$636K
from targeting offers by uplift instead of risk
03 · Models
Credit risk & loss lab
0.848
out-of-time AUC, ahead of the logistic benchmark
04 · Models
Forecasting at scale
45%
less forecast error than repeating last year
05 · Models
Next best action
+$195K
a month, estimated before launch
06 · AI engineering
Analytics copilot
9
governed metrics, each with an owner
07 · AI engineering
Credit policy assistant
89%
of questions find the right section first
08 · Products
Decision Story Studio
7
input formats, read in the browser
09 · Products
Decision platform
12
data tests run on every build
10 · Products
Pricing & demand copilot
7
stages, from framing to measured value
11 · Data & decisions
Causal impact lab
−0.43
difference-in-differences vs a true −0.40
12 · Data & decisions
Marketing mix optimizer
$1.9M
a year in lifetime value, same budget
13 · Data & decisions
Mortgage market intelligence
380+
U.S. metro markets scored
14 · Data & decisions
Mortgage product strategy
3
product lines compared
15 · Data & decisions
Lead prioritization
+38%
more conversions from the same team, on synthetic data
16 · AI engineering
Analyst agent
SQL
written against a read-only warehouse, replayed step by step
17 · AI engineering
LLM eval harness
30
test cases, graded by rules, with a release gate
18 · AI engineering
Document intelligence
40
synthetic invoices, extracted and checked by controls
Method notes
4
Tools
3