Nexus | Upcoming AI & ML strategy research

Planned instrument coverage

Backtesting across four instrument classes.

The Nexus roadmap covers Index, Equity, Futures and Options research. Each backtest uses the selected instrument, date range, capital, sizing, costs, and available historical data.

Index backtests

Test rules on supported Indian indices and index baskets. Combine indicators, price action, and multiple timeframes; compare portfolio contributions and risk.

Equity backtests

Research individual stocks and equity universes. Evaluate long/short rules where permitted, shared portfolio capital, position sizing, and symbol-level results.

Futures backtests

Test index and stock futures with dated contracts, expiry-aware rules, exchange lot sizes, costs, and explicit exposure assumptions.

Options backtests

Research index and stock options, buying and writing, single- and multi-leg positions, strikes, expiries, whole-lot sizing, combined risk, and leg-level reports.

Planned coverage is for supported Indian instruments and available data. It does not mean every exchange, contract, or historical period is included.

Upcoming product features

AI-assisted creation, machine learning tuning, analysis, and execution.

The following features describe the planned Nexus offering, from the first strategy idea to research, Paper/Live operation, and institutional administration.

01

AI-assisted strategy creation

Describe measurable entry, exit, timing, instrument, sizing, and risk rules in plain language. AI-assisted clarification turns your idea into a reviewed, saved strategy definition.

02

Visual & manual strategy authoring

Build a strategy through editable settings and rule controls as well as chat. Review its meaning, name it, and save a verified first version.

03

Versioned Blueprints

Save immutable strategy versions, inspect history and rule changes, revert through a new version, and continue a saved Blueprint in a fresh conversation.

04

Indicators, price action & multi-timeframe rules

Combine supported indicators and measurable price-action conditions with grouped rules and higher-timeframe confirmation. Planned custom formula and indicator support expands how explicit rules can be expressed.

05

Index, Equity, Futures & Options backtests

Backtest all four planned instrument classes: indices, equities, dated futures, and index/stock options. Preserve the selected market, contract, expiry, lot size, and data assumptions.

06

Single- & multi-leg options

Define buying or writing positions, straddles, strangles, spreads, condors, and expiry-aware structures. Track exact contracts, individual legs, combined risk, adjustments, and costs.

07

Deterministic backtesting

Execute reviewed rules against historical candles with explicit capital, sizing, timing, fees, and slippage. AI helps define and explain the strategy; the backtest executes its saved rules.

08

Basket & universe research

Test selected symbols or index-member baskets with shared portfolio capital. Inspect symbol contributions and any skipped instruments instead of relying on a single aggregate score.

09

AI analysis & result comparison

Ask AI to explain saved strategies, summarize backtests and machine learning optimization results, compare candidates, and highlight costs, drawdown, trade patterns, and risk assumptions. Review evidence-based candidate, baseline, or no-change recommendations.

10

Machine learning-assisted strategy tuning

Choose parameter ranges, objectives, constraints, and search budgets. Machine learning-guided adaptive search learns from trial outcomes to propose further candidates. Compare the baseline, ranked results, and out-of-sample evidence before accepting a new version.

11

Paper mode & broker-connected Live mode

Forward-test a reviewed strategy in Paper mode, then prepare a separately authorized Live deployment. Planned Dhan and Zerodha workflows cover orders, positions, P&L, and recovery.

12

Sizing, position management & risk controls

Define sizing, stops, targets, trailing exits, staged exits, and daily limits. The roadmap includes advanced position management, funds/margin checks, kill switches, and fill reconciliation.

13

Prompt Library & product assistance

Explore strategy-writing examples and indicator knowledge, ask product-operation questions, and analyze stored research through contextual assistance.

14

Private teams & account administration

Manage invitations, account access, user roles, saved research, and feature permissions. Keep private strategy/chat content separate from institutional operational and usage oversight.

15

Institutional white-label solution

A planned branded platform for licensed algo-provider institutions: organization branding, verified domains, client workspaces, roles, usage controls, and support. See institutional eligibility and regulatory responsibilities below.

16

Institutional marketplace

A planned marketplace for eligible licensed algo-provider institutions to publish and manage their authorized offerings, listing reviews, subscriptions, and commercial records. Publishing is restricted to institutions with the required legal permissions.

17

Plans, subscriptions & billing

Manage workspace plans, feature access, usage limits, Paper/Live capacity, upgrades, scheduled plan changes, billing, and institutional subscription records.

18

Training, support & platform operations

Use separate trader and administrator training, role-aware product help, support tickets and escalation, configurable AI/data providers, health monitoring, and usage reporting.

Upcoming · Machine learning for strategy research

Learn from parameter trials. Analyze the evidence.

The Nexus roadmap brings machine learning into strategy tuning through Bayesian adaptive parameter search. AI analysis combines the search outcomes with backtest, risk, and robustness evidence to help you review a candidate.

Adaptive parameter search

Bayesian/TPE search learns from completed trial results to propose the next parameter combinations within your chosen ranges, objectives, and constraints. Grid, Sobol sampling, and NSGA-II evolutionary search offer complementary ways to explore the same strategy.

Out-of-sample validation

Assess candidates through anchored walk-forward windows, a separate final holdout, parameter-neighbour stability, and overfitting diagnostics. Search results and validation evidence remain distinguishable so you can judge consistency.

AI analysis of ML search results

Compare the optimized candidates with the original baseline. Review AI explanations of trade-offs, risk, evidence strength, and cautions, including a recommendation to keep the baseline or choose no candidate when the evidence does not support a change.

You select the search scope and review any proposed strategy version. Optimization and analysis do not automatically rewrite your trading rules or authorize Paper/Live deployment.

Planned machine learning-assisted tuning & optimization

Explore parameters with machine learning. Test robustness.

Tune selected parameters of a saved strategy while keeping unselected rules, sizing, and execution assumptions intact. Compare reproducible candidates against the original baseline before accepting a new version.

Choose the search

Use Exact Grid for exhaustive combinations and seeded Sobol for balanced sampling. Planned Bayesian/TPE machine learning-guided search adapts to previous trials; NSGA-II provides evolutionary multi-objective exploration.

Validate out of sample

Use chronological research/confirmation/holdout splits or anchored walk-forward windows with purge and embargo. Keep final holdout evidence separate from search decisions.

Inspect overfitting & stability

Review Deflated Sharpe, probability-of-backtest-overfitting checks where inputs suffice, fold consistency, parameter-neighbour stability, and research-to-holdout degradation.

Planned Backtest → Tuning → Paper → Live workflow

Research and execution share your rules.

Define & verify

Describe or manually author the strategy, resolve clarification questions, and review its saved version.
Backtest & tune

Run historical tests, inspect trades and risk, explore parameters, and save the candidate you accept.
Validate in Paper

Forward-test the reviewed strategy in a simulated account before committing to an execution setup.
Prepare & operate Live

Review the broker connection, account permissions, capacity, funds, margin, and risk. Authorize deployment and monitor orders, positions, recovery, and exits.

Control stays with the trader

Explicit approval for run and deployment actions.
Saved versions link research and execution.
Separate Paper and Live capacity and account state.
Strategy sub-accounts distinguish Algo, other, and mixed activity.
Encrypted credentials, session controls, and retained execution audit.
Pause, exit, kill-switch, and recovery workflows.

Planned workspaces for research and institutions

A workspace that grows with your business.

Individual traders

Build and review your own measurable rules, compare research, tune parameters, and manage your Paper/Live workspace.

Research teams

Share an administered environment with defined roles, private artifacts, controlled feature access, training, and support.

Licensed algo-provider institutions

Explore planned white-label client workspaces and marketplace services for institutions with the applicable legal licences, registrations, and approvals. Institutional services follow the regulatory responsibilities below.

Planned connections & deployment

Configure the platform around your environment.

Nexus is planned as an independently deployable platform for customer-controlled research and administration. Broker-connected institutional execution must use hosting and connectivity arrangements permitted by the applicable exchange and broker requirements.

Provider choice, workspace branding, access roles, limits, and commercial arrangements are configured for your deployment. Infrastructure sizing, backup responsibilities, and optional AlgoMars management are agreed at onboarding.

Plan your installation

Institutional commercial workflows

Workspace subscriptions govern product access. Planned institutional marketplace listings and client subscriptions are administered within the organization’s authorized business scope.

AlgoMars provides the software and engineering services. Institutions remain responsible for their offerings, client relationships, regulatory permissions, and ongoing compliance.

Upcoming · For licensed algo-provider institutions

White-label platforms & an institutional marketplace.

Build a branded client platform and manage authorized algo offerings through Nexus. These planned solutions are intended for algo-provider institutions that hold the legal licences, registrations, empanelment, and approvals required for their business activities.

Your brand. Your client workspaces.

Use your logo, verified domain, role-based client access, child workspaces, subscriptions, usage reporting, training, and support escalation. Manage your institutional operation within its permitted service scope.

Marketplace for authorized institutions

Publish institution-owned offerings, manage listing review, customer subscriptions, billing, and commercial records. Seller onboarding and publishing are intended only for eligible institutions whose relevant permissions have been verified.

Regulatory responsibilities

Obtain the applicable authorizations before offering services.

Organizations must identify and obtain the licences, registrations, and approvals applicable to their proposed activities before serving clients, then maintain ongoing compliance. A Nexus software licence does not grant regulatory permission.

Confirm applicable SEBI registration and exchange algo-provider empanelment requirements for your business model.
Complete required broker arrangements and algo registration or approval processes before client execution.
For black-box algo offerings, meet the applicable Research Analyst registration and research-report requirements.
Follow applicable rules for broker-server hosting, order identification, risk controls, audits, and records.

Eligibility depends on the organization’s activities and the rules of the relevant regulator, exchange, and broker. Institutions should verify the current requirements with their compliance advisers before applying or launching an offering.

Official references: SEBI retail algorithmic trading framework · NSE retail algo guidance · NSE consolidated operational requirements

Discuss institutional requirements

Questions, answered

Know your platform.

Is Nexus available now?

Nexus is presented here as an upcoming product. These descriptions cover the planned feature set. Contact AlgoMars to discuss your research needs or institutional requirements; no launch date is promised here.

How is Nexus different from Options Lab?

Options Lab focuses on NIFTY and SENSEX index-options operations. Upcoming Nexus covers Index, Equity, Futures and Options backtests, AI-assisted strategy creation and analysis, machine learning-assisted parameter optimization, Paper/Live modes, and institutional white-label/marketplace solutions.

Does chat automatically place an order?

No. Strategy definition, backtests, tuning, and deployment use explicit review and action steps. A conversation is not permission to place broker orders.

Can Nexus test any trading idea?

Rules must be explicit, measurable, supported by the strategy engine, and backed by available market data. Nexus asks for clarification rather than silently dropping unsupported or incomplete conditions.

Who owns marketplace strategies?

The planned marketplace is for offerings published by eligible licensed algo-provider institutions. Each institution owns its content and must have the legal permissions required for the services it provides. AlgoMars supplies the software platform.

Can I operate Nexus under my own brand?

The planned institutional white-label solution supports your branding, domain, client workspaces, roles, plans, and usage controls. Commercial client-facing operation is intended only for institutions holding the applicable legal licences, registrations, empanelment, and approvals.

Do I need Options Lab to use Nexus?

No. Upcoming Nexus is designed for an independent installation and does not require Options Lab. Discuss the planned Nexus setup separately from an Options Lab deployment.

Does a Nexus licence authorize an institution to offer algo services?

No. A software licence is separate from regulatory authorization. Institutions must obtain and maintain the permissions applicable to their business model before providing client-facing algo or marketplace services.

How is machine learning used in Nexus?

Planned Bayesian/TPE search learns from historical parameter-trial outcomes to guide further tuning. AI analysis reviews those results alongside baseline, walk-forward, holdout, stability, and risk evidence. You decide whether to accept a candidate; each candidate is still evaluated through the saved strategy’s backtest rules.

Index. Equity. Futures. Options. Build, test, tune, and analyze with AI.