One decision, seven specialists.
Most trading tools are a single box that says buy or sell and asks you to trust it. ORO SAGE is built the opposite way: seven independent specialists that hand work to each other, and write down the reason at every step.
The short answer. ORO SAGE screens about 5,700 listed Indian stocks every trading morning, scores what survives, decides the action, then places and manages the order inside your own broker account. Stops and targets sit at the broker. Every decision is written to an append-only journal with its reason attached.
Four beats. Screen, decide, execute, audit.
Everything the system does sits inside one of these four. The seven roles underneath them are the detail, not a different story.
Scroll sideways to see the whole diagram.
Screen
Before the market opens, a fresh snapshot of about 5,700 listed names, 163 data columns each, lands and gets read. The Analyst scores every candidate against a fixed pattern set. The Screener sorts what survives into strategy books and separates what is tradeable today from what is only worth watching. Most names die here, which is the point.
Analyst · ScreenerDecide
The Optimizer takes the survivors and the positions already open, and decides the actual action: buy, sell, hold, trail the stop, or swap a weaker holding for a stronger one. Size comes from conviction, not from enthusiasm. Every buy candidate has to clear six gates before it is allowed to become an order.
OptimizerExecute
The order goes to the broker with the stop-loss and the target attached to it, so the protection sits at the exchange rather than in our software. Shortly after, the system reads the broker's own books and confirms what actually happened, at the price the shares actually fetched. If the two do not agree, it writes nothing and raises its hand.
ExecutionAudit
The Gatekeeper checks the other roles independently and can halt everything. The Reporter writes what happened into an append-only journal, reason attached, wins and losses in the same plain form. The Oracle reviews the week against stated goals and holds the guardrails the rest of the system runs inside.
Gatekeeper · Reporter · OracleThe seven roles.
Each role does one job and hands a clean, complete record to the next. If a role is unsure, it says so in writing and the trade waits. Four of them carry the trade forward. Three sit across the top and watch.
Analyst
Studies each stock against proven technical and fundamental patterns and assigns a conviction score it can defend. This is the gate everything downstream depends on. If a name does not earn the score, it never reaches the part of the system that can spend money.
Screener
Sorts scored stocks into the correct strategy lists, a short-term swing against a longer hold, and separates what is ready to trade from what is only worth watching. It is the difference between a good stock and a good stock right now, which are not the same thing.
Optimizer
The trading brain. On every position and every candidate it decides the action: buy, sell, hold, trail, or swap a weaker holding for a better one. It sets position size from conviction and tests each candidate against six gates before anything moves.
Execution
Places the order, sets the stop-loss and the target with the broker, and trails the stop as a trade moves in your favour. It owns the position book and keeps it exact to the paisa, and it only records what the broker has actually confirmed.
Gatekeeper
The built-in auditor. It independently checks that the other roles are doing what they claim, hunts for silent failures, verifies the data is fresh, and can halt everything. This is the role that exists so a problem is never quietly ignored, which is the way automated systems usually fail.
Reporter
Records everything exactly as it happened, with no spin and no summarising, in an append-only journal. Entries are added and never rewritten, so the record cannot be tidied up later to look smarter than it was. This is the complete trail behind every rupee.
Oracle
The governance layer. It reviews performance against stated goals every week, tracks whether the conviction scores were honest rather than merely lucky, and holds the guardrails the rest of the system must run inside. It is the only role allowed to change what the others are permitted to do.
Four roles carry the trade forward: study, sort, decide, execute. Three sit across the top and watch, record and govern. Separating the doing from the watching is the whole point. The parts that act cannot also be the parts that grade their own work.
Follow a single decision through all seven roles.
This is the same sequence every position goes through, on an illustrative stock we have called ACME LTD. The times are a normal trading day. The point is not the outcome, it is that every step leaves a record behind it.
Scroll sideways to see the whole diagram.
Before the open
The morning snapshot lands: about 5,700 listed names, 163 data columns each. ACME LTD is one row in it. The Analyst scores it 8.6 out of 10 against the pattern set, with the trend template complete and relative strength in the top decile of its cohort. The Screener files it as a swing candidate and marks it tradeable today rather than watch-only. Several thousand other names do not survive this step, and that is the step doing its job.
09:38 IST, the decision
The Optimizer runs ACME through the six gates, judging it against the account as it stands that morning. The regime reserve is intact, so cash is being held back and there is room to deploy. The sector sits under its ceiling. Conviction 8.6 sets the size ceiling this position cannot exceed. Entry ₹3,412.00, stop ₹3,268.50 which is 4.2 percent below entry, first target ₹3,742.00 and second target ₹4,094.00. Four shares, ₹13,648 committed, and the worst case if the stop fills is ₹574. That number exists before the trade does.
09:39 IST, the order
The Execution role places the order with the broker, stop-loss and target attached to it. From that second the protection lives at the exchange, not in our software. If our machine died right now, the stop would still be standing and the position would still be protected.
09:41 IST, the confirmation
The system reads the broker's own records, finds the fill, confirms the stop and the target are live at the exchange, and checks the position against what the account actually holds. Only now does the trade become true in our records. If any of it had failed to add up, the system would have written nothing and raised its hand instead. A late entry shows up the moment anyone reconciles. A wrong entry spreads quietly through profit and statistics wearing the same shape as a correct one.
09:41 IST, the journal entry
The Reporter writes the record: ticker, action, conviction score, entry, stop, targets, quantity, the money at risk, the gates the candidate passed, the market regime on the day, and the broker's own reference for the order. Every value carries its source and its age. The entry is appended and can never be rewritten.
- Conviction
- 8.6 / 10
- Entry
- ₹3,412.00
- Stop-loss
- ₹3,268.50
- Targets
- ₹3,742.00 / ₹4,094.00
- Quantity
- 4 shares (₹13,648)
- Worst case
- ₹574
- Gates passed
- 6 of 6
- State
- CONFIRMED at the broker
11:58 IST, the stop moves
ACME runs. Execution raises the stop from ₹3,268.50 to ₹3,415, which is above the entry price. From this moment the trade can no longer end in a loss. The new level is recorded in our book only after the broker confirms it accepted the change, because a stop that exists in our records and not at the broker is worse than no stop at all. It buys your peace of mind while protecting nothing.
14:22 IST, the first target
ACME touches ₹3,742. The broker fills the target itself, because that order has been sitting at the exchange since 09:39, and sells two of the four shares. Our system sees it on the next check against the broker's records and books the sale at the price the shares actually fetched rather than the price we hoped for. The remaining two shares ride toward ₹4,094 with the stop at ₹3,415 underneath them.
The weekly review
The Oracle reviews the week against stated goals: what the system said it would do, what it did, and where the two diverged. The calibration question is not "did it make money". It is "was the conviction score honest". A high-conviction trade that failed is a more useful journal entry than a low-conviction trade that happened to work, and it is the entry most services would quietly leave out.
Illustrative example with an illustrative stock. ACME LTD is not a real company and none of the above is a recommendation to buy or sell any security.
It earns real money the hard way
ORO SAGE was proved on paper first. Not one live order was sent in the system's entire build history until the whole machine had been tested end to end, and then it crossed to real money the only fair way we know: with the founder's own capital, live since August 2026.
Every rupee the system manages today belongs to the person who built it, and every decision on that money lands in the append-only journal with its reason attached, the losing ones in the same plain form as the rest. Other people's capital comes later, after the required registrations and after the live record, drawdowns included, has earned it. The reasoning behind that order of events is on the transparency page.
Common questions
How does the system decide what to buy?
Every candidate is scored against a fixed pattern set before anything else happens, and most names are eliminated at that step. What survives is sorted into a strategy book, then run through six gates before it is allowed to become an order. Conviction sets the position ceiling. Nothing enters on a feeling, and the reason is written down at the moment of the decision rather than reconstructed afterwards.
Can ORO SAGE guarantee returns?
No. No trading system can guarantee profits or prevent losses, and any service that implies otherwise is selling you something else. ORO SAGE is engineered to size carefully, to place its protection at the broker, and to stop itself when a number looks wrong. Capital is still at risk on every trade. The risk disclosure is the honest version and worth reading before anything else.
Who places the trades?
On Signals you place them yourself, on your own broker, at your own pace. On Autopilot ORO SAGE places them inside your own broker account for you. In both cases the capital stays in your custody, and the system has no path to withdraw money from the account. Autopilot opens after SEBI registration and broker algo-empanelment are in place. Both are described on the plans page.