Timeline
Timeline sells an integrated ecosystem to United Kingdom financial advice firms that spans three businesses sold as one proposition. The planning software began in 2018 as a retirement decumulation tool built on the academic sustainable withdrawal rate literature, letting an adviser illustrate historical worst case scenarios that other planning packages could not, and has since widened into lifelong cashflow modelling, a client factfind, digital letters of authority and an inheritance tax planner.
Alongside it sits an investment business comprising evidence based model portfolios and a multi asset fund range with portfolio analytics, which passed three billion pounds under management in 2025 and is distributed across fourteen platforms. A third arm is the company's own platform technology handling order execution and custody.
In May 2026 the company launched Pennee, a chat driven assistant reached inside the planning product, which prepares annual reviews, summarises meeting notes, drafts client communications and reports, structures financial plans from uploaded documents, builds cashflows, sends factfinds and assembles follow up workflows, drawing on client records, portfolio data and platform data already held in the ecosystem rather than starting from a blank prompt. It can join client meetings to record and summarise, extract action points and client vulnerability indicators, and flag life events.
The group operates through two companies with a boundary it publishes itself: the investing and platform services entity is authorised and regulated by the Financial Conduct Authority, and the planning software entity is not. Founded by Abraham Okusanya and headquartered in London.
Capability Axes
Capability grades
15 of 15 axes rated · 4 graded A or B
The established platform reading applies and the arithmetic behind it is unambiguous. Everything sold before May 2026 survives the removal of the assistant intact: the withdrawal rate and cashflow engine that was the whole product for eight years, the factfind, the digital letters of authority, the inheritance tax planner, the model portfolios and multi asset funds carrying over three billion pounds, the portfolio analytics, and the platform handling order execution and custody.
The assistant is one named item inside the planning menu, three months old at the time of assessment. The vendor's own framing is that this is artificial intelligence embedded in an integrated ecosystem rather than bolted onto a disconnected stack, and that framing is fair as a description of how it is wired, but it describes where the models sit rather than how much of the product they carry. Strip them and an award winning planning package and a discretionary investment business both remain fully operational.
The boundary is published, categorical and covers every write path, which puts it well above the norm on this axis. The assistant is stated never to send, submit, change or trigger anything independently, every output is reviewed by the adviser before it goes anywhere, and every action requires explicit approval. Four verb classes named in one sentence is a more complete statement of the limit than most vendors manage.
Held below the top grade because it is asserted as a design principle and nothing describes the mechanism behind it: searched for an administrator control over the gate, a record of what was generated against what was approved, an audit trail a compliance reviewer could examine, or any way a firm could verify the boundary holds rather than take it on trust, and located none. A stated principle and an enforced configurable control are different things to a firm that has to evidence its supervision.
Searched for an accuracy figure, an error rate, a benchmark, a validation method, a grounding or citation guarantee tying a generated statement back to the record it came from, a drift statement or a revalidation cadence, and located none. The unmeasured risk is unusually concrete for this vendor because of what the assistant produces.
Structuring a financial plan from uploaded documents and generating a client report that reflects both a meeting discussion and portfolio analytics means assembling figures that end up in a suitability file, and a plan drafted from a misread document is wrong in a way the reviewing adviser is poorly placed to catch, since the draft looks complete and the source document is no longer in front of them.
The adoption evidence is substantial and it is entirely about the investment business rather than the software or the assistant. Assets under management passing one billion pounds within roughly two years of launch and three billion pounds by 2025 is money actually placed by advice firms, which is a harder signal than a testimonial, and the company runs its own adviser conference and publishes an awards page.
What is absent is any evidence about the thing this index is grading: searched for a named advice firm reporting an outcome from the planning software, any measured time saving, any adoption figure for the assistant, or any customer account of using it, and located none. The assistant launched in May 2026 and the vendor states advisers are already using it without naming one. Queued check: the awards page and any case study library.
Recorded as an unverified absence. Searched the assistant announcement, the product pages and the launch coverage for any statement on whether recorded meetings, transcripts, generated drafts or client records are retained, used to improve the service, or used to train, tune or evaluate any model, and located none in either direction.
The question has an edge specific to this vendor's structure, because the same organisation holds the client records, the portfolio positions and the platform custody data, so material flowing through the assistant is drawn from a wider pool than a standalone note taking tool would ever touch, and nothing published describes the boundaries around that pool.
Recorded as an unverified absence. A privacy policy is linked from the site footer and was not opened, so this row reflects material not read rather than material established to be missing, and that limitation is stated deliberately.
What the assistant does raises the questions a policy would need to answer: it joins client meetings to record and summarise them, extracts indicators of client vulnerability, and draws on client records, portfolio holdings and platform data in the same operation. Vulnerability indicators about an identifiable individual are among the more sensitive categories a firm holds. Searched the product and announcement material for retention, deletion or processing terms and located none there. Queued check: the linked privacy policy and any processing terms offered to advice firms.
Recorded as an unverified absence rather than an evidenced one. Searched product pages, the assistant announcement and the site navigation for an enumerated attestation, a named information security standard, an audit report or a dedicated trust page, and located none, and no security page was found to open.
The consideration is sharper than for a pure software supplier because the group operates a custody platform and holds client assets, which brings operational resilience and outsourcing obligations to the authorised entity, and because the assistant now processes recorded client meetings. Those obligations sit with the regulated company rather than the software company, and nothing published bridges the two. Queued check: a trust or security page and the regulated entity's published disclosures.
This is the sixth application of the principle that supervision of one part of a group does not read across to software sold by another, and it is the first where the vendor draws the line itself in its own footer rather than the line having to be inferred. Investing and platform services are provided by one company, named with its registration number and its authorisation and reference number at the conduct regulator.
Planning software and tools are provided by a second, separately numbered company, and the published statement that this entity is unregulated is set in bold. The assistant sits inside the planning entity. So a genuine authorisation exists within the group, and it covers discretionary investment management and custody rather than the generative tooling. Recording the honesty as a positive shape even though it earns no grade here, because most vendors in this index would have let a reader assume the authorisation covered everything.
One capability here sits closer to the purpose of this axis than almost anything else in the pocket, and it is undiscussed. The assistant automatically extracts indicators of client vulnerability from meeting conversations, which under United Kingdom consumer protection expectations is a determination with direct consequences: a client correctly identified receives adapted service, and a client the system fails to identify does not.
Detection of that kind depends on how people express difficulty, which varies with age, health, language, culture and how willing someone is to disclose, so uneven performance across client groups produces uneven protection for exactly the people the requirement exists to protect. Searched for any evaluation of vulnerability detection across client types, any account of what the system treats as an indicator, or any fairness testing, and located none.
Searched for a warranty, an accuracy commitment, a correction obligation or any allocation of responsibility between vendor and advice firm, and located none. The oversight design carries the allocation implicitly: a boundary under which the assistant never acts without explicit adviser approval is a real control and simultaneously places every consequence of a wrong draft on the person who approved it, and the vendor states plainly that the firm remains in control at all times.
The party with no route at all is the client, who has no described means to learn that a vulnerability indicator was recorded about them by a system, to see it, or to contest it, and that record then shapes the service they receive.
Searched the assistant page, the launch announcement, the trade coverage of the launch and the product navigation for a model provider, a model family, a version, an inference host or a subprocessor list, and located none. The vendor's positioning makes the silence conspicuous rather than routine: its argument against competing tools is that they are generic and disconnected while this one is connected to real client data, an argument entirely about where the data comes from and never about where the reasoning happens. A reader is told precisely which internal systems feed the assistant and nothing about whose model consumes them.
The connectedness is real and it points inward, which is the distinction that decides this grade. The assistant's stated advantage is that it reaches across planning, platform, portfolios, reports, workflows and client records without the adviser moving between systems, and that is genuine depth because the vendor owns every one of those surfaces.
It is also the limit: an advice firm running a different back office gets nothing from it, because the connections are to this vendor's own stack rather than to the estate a firm already has. Outward reach exists elsewhere in the business, through model portfolio distribution across fourteen platforms, digital letters of authority that connect to providers, and an account aggregation integration. Searched for a programmatic interface or named connections into third party client relationship systems and located none.
Recorded as an unverified absence. The products are reached through a browser with a hosted authentication provider, and nothing about topology or location is published in the material located. Searched for hosting arrangements, a storage region, a residency commitment, or any statement of where recorded client meetings and generated content are processed, and located none.
For a vendor selling exclusively into United Kingdom advice firms handling United Kingdom client data, a residency statement is the disclosure a buyer would expect to find first and it was not present on any page opened.
Searched for a rate, a tier, a per adviser figure, a fund or portfolio charge, or any statement of whether the assistant carries its own line or is included in a planning subscription, and located none. The commercial question is more interesting here than for a single product vendor and is entirely unanswered: because the company sells software, discretionary portfolios and a custody platform to the same buyer, a firm cannot establish from published material whether software pricing moves depending on whether that firm also uses the portfolios, which is the arrangement most worth understanding before adopting any part of it. Queued check: the vendor's own pricing material and the fund and portfolio charge documents.
Breadth here comes from product range rather than from buyer variety, which is an unusual shape for this axis. The buyer is a United Kingdom financial advice firm and nothing else, with no bank, insurer or asset manager segment addressed, and the international footprint the company described in earlier years is not visible in current positioning.
Against that single buyer type the company sells planning software, discretionary model portfolios, a multi asset fund range, portfolio analytics and its own custody platform, so an advice firm can take one component or the entire stack, and the portfolios reach firms that use none of the software through distribution across fourteen third party platforms. Held below the top grade on the single jurisdiction and the single professional buyer.
Alternatives to Timeline
The closest documented capability profiles to Timeline in the same categories, ordered by similarity across the same fifteen axes the index grades every vendor on. Closest documented profile, not a claim that either product does the same job. No vendor pays for placement.
Documents Model Risk Management and Transparency where Timeline does not
Stronger documented coverage on Autonomy and Oversight Model
Stronger documented coverage on Institution and Segment Coverage
A lighter documented profile than Timeline
Documents AI Centrality and GLBA and Data Privacy Posture where Timeline does not
Documents Deployment Model and Data Residency where Timeline does not
Similarity is computed axis by axis from published grades, not from a composite score. The index does not aggregate grades into a total. See the fifteen axes and the methodology.
Pricing
Vendor-published figures are labeled as such. Figures labeled “Estimated” are derived from third-party sources and have not been confirmed by the vendor.
No pricing data has been verified for this vendor. Pricing information will be published here once confirmed through vendor disclosure or third-party estimation.