✦The desk, without the barrier· start on paper
Investing tools you can actually use.
Research, simulate, and act on a thesis — even if you have never written a backtest. Your Agent does the heavy lifting and proposes. You decide.
Alphonce is an Ochieng & Co product.
Research · AI chat
What's driving NVDA's margin expansion this quarter?
Data-center revenue grew 154% YoY, with Hopper shipments
10-Q · Q3'26Earnings call
Portfolio
Total value
$284,310
▲ $2,108 · 0.74% today
Equities 64%
Bonds 22%
Alts 9%
Cash 5%
Backtest · 2019 – 2026
Momentum + Quality Tilt
vs SPY benchmark · monthly rebal.
CAGR
+18.4%
Sharpe
1.42
Max DD
−12%
Brokers
Connect your account
T
I
A
C
F+8
Order filled
42 shares · $24,810
Attribution logged
Attribution logged
Connect the brokers you already use.
Live today, more on the way — your assets stay put.
Live today, more on the way — your assets stay put.
One question.
The whole loop.
The whole loop.
Type a thesis in plain English — the Agent researches, simulates, and proposes. You approve.
$ run thesis
ETA 1.8sThe full loop
Stop stitching together six tools.
Screeners, research, backtesting, brokers, analytics — each one a separate product, each one a barrier. Alphonce folds the loop into one place you can actually sit down and use.
01
Ask, don't query.
Plain-English research across filings, transcripts, and live data — with sources, not a fluent guess.
02
Build without a quant seat.
Visual studio or sandbox. Simulate with walk-forward and stress tests before anyone spends a pound.
03
Act where you already trade.
The Agent proposes. You approve. Orders go through the broker you already have.
How it works
Five steps.
Anyone can start.
Each stage connects to the next — no exports, no glue code, no second job. The barrier was the stack. The stack is now one path.
01
Connect
Link a broker you already have. Read-only or trade-enabled — no transfer, no new account.
02
Ask
A question in plain English. The engine computes; the Agent explains, with sources.
03
Build
Assemble a thesis visually or in the sandbox. You do not need to be a quant to start.
04
Simulate
Backtest, walk-forward, Monte Carlo, stress — so you can see the idea before you fund it.
05
Approve
The Agent proposes. Paper, dry-run, or live — you release the capital. Every fill keeps its thesis.
Capabilities
What's inside
the platform.
The tools a desk would build — research, simulate, propose — without needing the desk. Ready the first time you sit down.
AI research
Grounded answers from
filings, transcripts, live data.
filings, transcripts, live data.
Every claim links back to a primary source. No hallucinated numbers.
10-K · NVDA · 2026Q3 earnings callFRED · CPI seriesBloomberg terminalReuters feed+47 sources
Live portfolio.
Track performance, attribution, risk — across every account, every broker.
YTD
+14.2%
▲ vs SPY +3.1%
Backtest in seconds.
Point-in-time data, survivorship-bias free, with full transaction costs.
Sharpe
1.42
Max DD
−12%
Code when you want to.
No-code when you don't.
No-code when you don't.
A visual builder that compiles to real Python. Eject anytime.
# Alphonce → Python
from alphonce import Strategy, factors
strat = Strategy("momentum-quality")
strat.add(factors.momentum(lookback="6M"))
strat.add(factors.quality(metric="ROIC"))
strat.backtest(start="2019-01").summary()
Pricing
Serious tools.
A way in.
Start free. Upgrade when you are ready to take a proposal live. Cancel anytime — no lock-in, no data held hostage.
Starter
$0/forever
For researchers exploring the platform.
AI research (50 queries/mo)
1 connected broker, read-only
Paper trading
Basic backtesting
Community support
Most popular
Pro
$49/month
For serious retail investors and enthusiasts.
Unlimited AI research + citations
3 brokers, trade-enabled
Full backtest engine · 20Y history
Strategy builder + Python eject
Priority email support
Quant
$199/month
For power users running systematic strategies.
Everything in Pro
Unlimited brokers & sub-accounts
Tick-level data · options, futures
MCP API · run your own models
SOC 2 audit logs, SSO
“I cancelled Bloomberg, Koyfin, and three other subscriptions. Alphonce replaced all of them — and the research feels like working with a junior analyst who never sleeps.”
“Backtests that used to take me an afternoon in Python are now a 20-second conversation.”
Priya Ramanathan
Systematic trader
Systematic trader
“The attribution layer is what got me. I finally know which theses actually made money.”
Jonas Berger
Family office · Zurich
Family office · Zurich
Built for people who
want to do the work.
Connect a broker, ask a question, see a proposal — without a terminal, a quant, or a second job.