AI Portfolio Management Agent

Constructing an optimal allocation is weeks of modelling assumptions, testing scenarios, and weighing constraints while the committee waits. An AI Portfolio Management Agent runs that construction and optimization against your mandate. It builds the options; the PM decides. Tecla builds it inside your environment and runs it, wired into the market data, optimization models, and mandate constraints your team already uses.

The shift underway

Portfolio construction is shifting from static, mean-variance models to AI that tests thousands of scenarios, while managers deliberately keep the allocation decision human. The numbers show the shift and the line.

70%

of buy-side firms are actively using AI to support their front office.

Source: SimCorp InvestOps 2026

55%

have integrated AI into at least one investment process, keeping it a partner, not the decider.

Source: Mercer 2026

91%

of asset managers plan to increase their use of AI in the next 12 months.

Source:

How it works

How the agent works

Portfolio construction is a modelling cycle, assumptions, optimization, scenarios, constraints, that eats weeks of a quant team's time before the committee sees options. The agent runs that cycle and prepares ranked, stress-tested allocations for the manager to choose from.

01

Assumptions

Refreshes the capital market assumptions, returns, risk, and correlations across asset classes, that every optimization depends on, from current data.

Reads the room
02

Optimization

Constructs candidate allocations across competing methods, mean-variance, risk parity, and others, rather than a single model's answer.

Builds the allocations
03

Constraint Check

Applies the mandate and IPS constraints, limits, exclusions, liquidity, and turnover, so every candidate is investable within the rules.

Honors the mandate
04

Scenario & Stress

Stress-tests each allocation across market scenarios and measures drawdown, CVaR, and factor exposures, so the risks are visible before the choice.

Tests the risk
05

Candidate Compare

Ranks the candidate allocations with their trade-offs laid out, ready for the portfolio manager to weigh and decide.

Prepares the options

Human in the loop

Every candidate allocation carries the assumptions, constraints, and stress results behind it, so the PM weighs real trade-offs rather than a black-box number. The agent constructs and tests; the allocation decision stays with the portfolio manager.

Scoped to you

This is a typical portfolio management build. The exact parts are scoped to how your firm runs: your optimization methods, the capital market assumptions, the mandate and IPS constraints, and your risk framework.

See what this would look like built around your portfolio operation.
How it reaches production

Built, run, and owned, one phase at a time

An optimizer that looks elegant in a demo can still produce allocations that ignore a real mandate constraint. A scoped brief, a named team, and someone accountable at each stage are what make this one dependable on real capital.

01
Sprint

Scope, fixed price

We map your optimization methods, the assumptions and data, the mandate constraints, and your risk measures, then return a plan and a price before you commit.

AI Solutions Lead
02
Build

Built into your environment

Senior engineers assemble the agent inside your environment, wired into the market data, optimization models, and mandate systems your team already works in.

AI Systems Lead
03
Run

Monitored and improved

The agent adapts as your methods and assumptions change, and we report against the metrics set at scope: allocations constructed, scenarios tested, time to committee-ready options.

Systems Lead + Engineers
04
Own

Yours, improving

The system is yours to keep, and it stays aligned to your methods and mandates as they change rather than hard-coding one optimization anyone could buy.

Your team

A build starts with a fixed-price Tecla Sprint that scopes and prices it before you commit.

‍ Book a scoping call
How it holds up

Trustworthy enough to allocate real capital

An allocation the PM cannot interrogate is worse than none, because real capital and a mandate ride on it. What makes this one dependable is that every candidate shows its assumptions and risks, and the decision stays with the manager.

Reviewed by trajectory

Every allocation shows its inputs

Each candidate allocation carries its capital market assumptions, constraints, and stress results, so a PM can interrogate why it looks the way it does before choosing. The agent computes; it does not hide the math.

Human in the loop

It constructs, the PM allocates

The agent constructs and stress-tests the candidates, but the allocation decision and the judgment behind it stay with the portfolio manager. It prepares the options; it does not pick.

Tuned over time

It stays inside the mandate

As mandates, IPS constraints, and methods change, the agent is kept current, so every allocation it builds is investable within the rules that apply now.

Start the build

Build the allocations, keep the decision

A scoping call maps your meeting workflow and returns what the agent would handle, how success gets measured, and the fixed-price Sprint to build it.

Fixed-price AI Systems Sprint. No commitment until scope is confirmed.

What PMs ask about construction

What is an AI portfolio management agent?

An AI portfolio management agent constructs and optimizes candidate allocations against your mandate: it refreshes capital market assumptions, runs the optimization across methods, applies constraints, stress-tests the results, and ranks the options. Tecla builds and operates the portfolio management ai agent inside your environment, so the PM chooses from prepared, stress-tested allocations instead of waiting weeks on the model.

How is an AI portfolio management agent different from an optimizer or an OMS?

An optimizer gives one model's answer and an OMS executes trades; the agent runs the whole construction cycle. It sets assumptions, builds allocations across competing methods, applies the mandate, and stress-tests them. Tecla builds this portfolio construction ai as a layer over the models and data you already use, not a single black-box optimizer.

Does the portfolio management agent decide the allocation?

No. The agent constructs and stress-tests the candidates, but the allocation decision and the judgment stay with the portfolio manager. Institutional portfolio ai like this is a partner, not the decider, which matches how managers deliberately keep the allocation call human.

How does the agent handle mandate and IPS constraints?

It applies your limits, exclusions, liquidity, and turnover constraints to every candidate, so each allocation is investable within the rules. The Investment Policy Statement that guides your managers governs the agent too, which is what keeps portfolio construction ai inside the mandate.

Can the PM see the reasoning behind each allocation?

Yes. Each candidate carries its capital market assumptions, constraints, and stress results, so a PM can interrogate the math rather than trust a number. Portfolio management automation that hides its reasoning is unusable at the institutional level, so the agent shows its inputs.

What does Tecla's portfolio management build cover?

Tecla scopes the build to your firm: your optimization methods, the capital market assumptions, the mandate and IPS constraints, and your risk framework. Built as institutional portfolio ai, it runs inside your environment and Tecla operates it, updated as your methods and mandates change.

How is our data and methodology kept secure?

The agent runs in your environment, integrated with your systems, under your own access rules. Your models, assumptions, and portfolio data stay within the boundaries your firm already answers to, and you own and control the system Tecla operates.

How long does a portfolio management build take?

A build starts with a fixed-price Tecla Sprint that scopes and prices it before you commit. The Sprint maps your methods, assumptions, and constraints, so the portfolio management ai agent that follows fits how your firm actually constructs portfolios.

We have a quant and PM team already. Where does the agent fit?

The agent runs the assumption-setting, optimization, and stress-testing so the team spends its time on the allocation decision and the mandate, not the modelling cycle. Tecla builds and operates the portfolio management automation alongside them, with every candidate showing its inputs.

How do we get started with Tecla?

A scoping call maps your construction workflow and returns what the agent would build and test, how success is measured, and the fixed-price Sprint cost. Book a scoping call with Tecla to start.

Have any questions?

Schedule a call to discuss in more detail.

Book a call