Using Agents for Long-Horizon Financial Work — Aman Singh, Arca

“We're actually not a SaaS company. We are, um, we are the firm ourselves. Agents prompt people more than people prompt agents.”

Claim: Arca describes itself as a firm rather than a SaaS company, where agents prompt people more often than people prompt agents.

“And it is a proactive kind of, um, service that is running for them twenty four/seven and-”

Claim: Arca's service operates proactively for clients around the clock.

“Yeah, for sure. Um, thanks for having me. Um, my role at Arca is I lead the AI platform team, so we're building the underlying sort of, um, harness and our capabilities to kind of run the business with AI, which is our ultimate goal.”

Claim: Aman Singh leads Arca's AI platform team, which is developing the infrastructure and capabilities for Arca to run its business with AI, the company's ultimate goal.

“Um, in terms of the services that we provide to our clients, um, with-- This is a broad spectrum of financial services. So that includes, um, sort of financial advice, financial planning for, um, for their assets, um, and sort of, um, uh, tax planning and insurance and estate planning as, as people get wealthier.”

Claim: Arca offers clients financial advice and planning, tax planning, insurance, and estate planning.

“Um, and there's lots of complexities with all the different partners, et cetera, right? So, uh, we're gonna help you think through that. Um, we also have clients that are, you know, uh, really large technology executives that, uh-”

Claim: Arca helps clients navigate the complexities of coordinating the different partners involved in their financial affairs.

“... have, uh, are trying to figure out a real plan for retiring, retiring early, right?”

Claim: Arca has technology-executive clients seeking a concrete plan for early retirement.

“Um, and also sma- small families that, um, uh, that are trying to buy their first home, for example.”

Claim: Arca has small-family clients seeking to purchase their first home.

“Yeah. So, um, the role of AI at Arca is not to replace the advisor, it is to replace all of the work around the advisor.”

Claim: Arca uses AI to automate work surrounding financial advisors rather than replace the advisors themselves.

“So our clients don't really interface with AI directly.”

Claim: Arca's clients generally do not interact directly with its AI systems.

“All they experience is a level of service that has not been, been, um, been possible in this industry before- ... delivered through the empathy and care of a human advisor. Um, so we're building the machinery behind, um, every client-”

Claim: Arca says it delivers a previously unattainable level of industry service through human advisors' empathy and care.

“... delivered through the empathy and care of a human advisor. Um, so we're building the machinery behind, um, every client- ... that's constantly thinking about, um, you know, um, "What's, what is the market doing, and how does it impact Richard?"”

Claim: Arca is building systems for each client that continuously monitor market developments and assess their effects on that client.

“You know? Um, what i- what, what else is going on in your life? Like, what are you-- what are the s- the, the textures of your life that impact your finances? Um, and, uh, we're able to be way, way more proactive, way more responsive, um, and just deliver a better, uh, kind of, uh, level of client service using AI that does all of the, the back office stuff.”

Claim: Arca's client-support systems take account of life circumstances that may affect a client's finances.

“You know? Um, what i- what, what else is going on in your life? Like, what are you-- what are the s- the, the textures of your life that impact your finances? Um, and, uh, we're able to be way, way more proactive, way more responsive, um, and just deliver a better, uh, kind of, uh, level of client service using AI that does all of the, the back office stuff.”

Claim: Arca says AI handling back-office work enables it to provide more proactive, responsive, and better client service.

“Yeah. Um, so today we're using, uh, the Frontier, uh, models from, um, from, uh, the Frontier Labs, like Anthropic and, and OpenAI.”

Claim: Arca currently uses frontier AI models from Anthropic and OpenAI.

“Um, Claude is kind of our primary, um, um, model, has been.”

Claim: Claude has been Arca's primary AI model.

“But, um, we've gone through several iterations of models over the years. Um, we used to be in Gemini way back when. Um, and then OpenAI, and then, and then Anthropic. And what's guiding all of our decisions are, um, our benchmarks that we have internally for, um, all of the various, um, capabilities of our general harness, uh, that it has to express in, in running our business.”

Claim: Over the years, Arca has changed models several times, moving from Gemini to OpenAI and then Anthropic.

“Um, and then OpenAI, and then, and then Anthropic. And what's guiding all of our decisions are, um, our benchmarks that we have internally for, um, all of the various, um, capabilities of our general harness, uh, that it has to express in, in running our business.”

Claim: Arca chooses models using internal benchmarks of the capabilities its general-purpose harness needs to run the business.

“Um, and so, uh, just recently, for example, um, despite the external benchmarks claiming, um, great things for Luna, it didn't really quite work out for us as, as well as we had expected.”

Claim: Luna performed less well for Arca than expected, despite favorable external benchmark claims.

“It was way too slow or, um, and it was costing a lot.”

Claim: Arca found Luna too slow and expensive for its needs.

“And so, um, our internal benchmarks kind of guide our model decisions. And, uh, as a principle, we, we want to be, um, model agnostic- ... um, and, and sort of not, um, uh, you know, um, take a dependency on, on a, a certain provider or a certain harness from one of these labs, and kind of keep that, that optionality open for us.”

Claim: Arca uses internal benchmarks to choose models and aims to remain model-agnostic, avoiding dependence on any one provider or lab harness so it can retain flexibility.

“No. We have, uh, we have our, our own- ... uh, kind of built internally that is mo- model agnostic.”

Claim: Arca's internally built harness is model-agnostic.

“Well, yes. So we have, um, certain tiers of models based on different complexities of, of tasks.”

Claim: Arca assigns model tiers according to the complexity of each task.

“So a task like assisting with a tax plan or doing a tax analysis would get our, like, heaviest tier of models.”

Claim: Arca assigns its most capable model tier to tasks such as tax planning assistance and tax analysis.

“Inside of that tier, it is a little bit flexible, so you could use, um, so like an Opus tier model or a Sonnet tier model, uh, or a, a, a heavy GPT model.”

Claim: Arca's heaviest model tier can include an Opus-tier model, a Sonnet-tier model, or a heavy GPT model.

“Um, but, um, based on, uh, there's a lot of, like, background activity that our agents are participating in, right? That is using the, a lower tier of model- ... um, based on the complexity of the task. So that's kind of how we, how we make decisions-”

Claim: Arca assigns lower-tier models to background agent activity according to task complexity.

“Not really. Um, they're, they're, they're not aware of- ... uh, of that happening. We've abstracted this away- ... uh, from them entirely. They kind of just interact with our product, um, as they would. Um, and those decisions are made in- internally, uh, by us.”

Claim: Arca handles model-selection decisions internally and hides that process from financial advisors, who are not aware of it.

“Right. Um, so just like in the talk, this is a thing that will span 10 days. Um, and so agents are ultimately, um, responsible for delivering that outcome.”

Claim: In the example from the talk, agents are responsible for delivering the outcome of a process spanning 10 days.

“But there are, are, are certain parts that they do and certain parts that they do not.”

Claim: Agents handle some parts of the 10-day process, while other parts are not performed by agents.

“So in this example, um, when, uh, when the client emails us, uh, with this whole tax bill, there's a, there's a piece of the analysis of, like, where is this client's money allocated today, and what is the most optimal way, um, to liquidate some of that to, to produce $100,000? That is best suited for agents- ... to take on as an analysis task.”

Claim: Agents are best suited to analyze a client's current asset allocation and determine how to liquidate assets to raise $100,000 for a tax bill.

“So the analysis part is fully automated, but, um, they're unable to actually execute, uh, any sort of money movement or liquidation automatically.”

Claim: The company's analysis is fully automated, but its agents cannot automatically move money or liquidate assets.

“We have a very deep integration with our custodian, uh, that actually gives us APIs to be able to do that.”

Claim: The company has a deep integration with its custodian that provides APIs capable of moving money and liquidating assets.

“Um, but we have strict guardrails around human in the loop for things that would either move money or liquidate assets or even directly communicate with our, with our end clients, uh, because there's a level of service that, that, uh, service, service and quality that we expect.”

Claim: The company requires human involvement in actions that move money, liquidate assets, or communicate directly with clients, in order to maintain its expected service and quality.

“Um, so all of the, like, internal analysis will be done by agents, but the, the-- they will also pull in the right person to be able to execute the trade when needed.”

Claim: The company's agents perform internal analysis and involve the appropriate human to execute trades when needed.

“And they'll pr- give a, give the person a plan that they, they're able to review and kind of, like, adjust, um, and yeah, go from there.”

Claim: The company's agents provide a plan for a human to review and adjust.

“Yeah. Um, so this is an interesting, interesting one to think about is we have tons of these signals like I described in the talk. Um, some are, for example, you can imagine there's an email thread that is receiving responses, and there's chatter happening, right?”

Claim: The company tracks multiple signals about client situations, including active email threads with incoming responses and discussion.

“You can imagine that, that any ad- additional response to that goes to the same agent that's kind of, like, reading that, that email thread and reacting to it, right?”

Claim: The company's system can route new replies in an email thread to the same agent already handling that thread.

“But a, a fresh signal of like a, like a phone call that was made in the middle, um, that is a brand-new signal that isn't attached with any other ongoing process.”

Claim: A phone call received during an ongoing client process is treated as a new signal unlinked to another ongoing process.

“And so there's, like, a fresh agent that wakes up to figure out what this means, right?”

Claim: When a new, unlinked signal arrives, the company's system starts a fresh agent to interpret it.

“And the way they, way they understand what is happening right now is our internal memory systems.”

Claim: The company's agents use internal memory systems to understand the current situation.

“So we have, um, we have this, this persistence of, for every client, um, what is happening right now, what are we trying to service, who owns what, what has happened so far- ... what are we tracking for next, right? What, what needs to happen immediately in the next couple of days? What are we tracking for the long term, right?”

Claim: The company's memory system maintains per-client records of the current situation, service objectives, ownership, past events, and upcoming items to track.

“... what are we tracking for next, right? What, what needs to happen immediately in the next couple of days? What are we tracking for the long term, right?”

Claim: The company's per-client memory system tracks both immediate needs over the next couple of days and longer-term items.

“Which individual is responsible for which action? All of this is being tracked by our memory systems-”

Claim: The company's memory system records which person is responsible for each action.

“... that our agents contribute to. But also, there is a, a kind of really sort of dreaming process that reflects on the state of the world as it relates to Richard, the client- ... um, and is able to kind of like update these memory systems to, to persist long-horizon kind of tasks.”

Claim: The company's agents contribute to client memory systems, while a separate dreaming process reflects on the client's situation and updates those memories to support long-horizon tasks.

“So that agent that received a new signal quickly ramps up on what, uh, the situation is with the client, and it'll know, okay, we're, we're actually executing a big tax bill.”

Claim: An agent receiving a new signal quickly uses the client's stored context to understand the situation, including that a large tax bill is being handled.

“... and we'll be able to see, okay, my signal relates to that tax bill. So let me kind of contribute to the, to, to the state of that world.”

Claim: An agent can connect a new signal to a client's ongoing tax-bill process and update the associated client context.

“Yeah. Um, so all of our safety restrictions are, are fully deterministic, and they, they take the decision-making off the, the agent's hands.”

Claim: The company's safety restrictions are fully deterministic and take safety-related decisions out of the agents' hands.

“So for example, um, certain tools let the, let the agent kind of execute the tool itself. So if it's like a simple record-keeping kind of tool-”

Claim: The company's agents can directly execute simple record-keeping tools.

“... um, it will execute it. We also, with every time a tool is called that writes data, we have, um, uh, LLM-based verification running on, uh, on that action, um, these, like, fast, uh, LLM calls that, that analyze every single write to our database.”

Claim: The company uses fast LLM-based verification to analyze every tool action that writes to its database.

“Um, so that's another layer of verification we're running. And then, uh, we also ingest a lot of, like, external information, so there's emails coming in, documents coming in, phone calls being made- ... which are targets for prompt injection. So we run, um, sort of verifications at the, at, at the edges that detect presence of certain, um, you know, flavors of prompt injection.”

Claim: Because ingested emails, documents, and phone-call information can carry prompt injections, the company runs edge checks to detect certain forms of prompt injection.

“Um, so that's another layer of verification we're running. And then, uh, we also ingest a lot of, like, external information, so there's emails coming in, documents coming in, phone calls being made-”

Claim: The company's system ingests external information, including emails, documents, and phone-call information.

“And then anytime there is a, um, there's a case where, uh, data leaves our system, for example, in like a, a web search in, in the most obvious case, but also if we're sending a calendar invite to Richard, right? That's an external kind of, uh, data where we have both deterministic checks on presence of certain information that we know is PII-”

Claim: The company applies deterministic checks for known personally identifiable information when data is sent outside its system, including in web searches and calendar invitations.

“... uh, but also a consensus of small models that, that ensure that there isn't any kind of data exfil being attempted.”

Claim: For outbound data, the company uses consensus among small models to check for attempted data exfiltration.

“Um, so we have tons of different, different kind of guardrails built into the harness, uh, with the principle that we want to kind of contain them at the edges of capability, uh- ... and not earlier than that. Yes.”

Claim: The company's harness includes multiple guardrails designed to constrain agents at the boundaries of their capabilities, rather than restricting them earlier.

“Yeah. The, the industry cares about outcomes, but, um, compliance cares about auditability.”

Claim: In the financial industry, business stakeholders focus on outcomes, whereas compliance focuses on auditability.

“Right? Um, and so, um, our, um, our system is, is, is such that auditability is by default.”

Claim: Auditability is built into the company's system by default.

“Um, we have every single action that, that an agent takes produces a change log of what changed- ... um, who changed it, from which part of the system, and when.”

Claim: Each agent action generates a change log that records the change, its actor and system source, and its time.

“And this applies not just to agents, by the way. It applies to human actors as well.”

Claim: The company's change logging covers human actions as well as agent actions.

“So if you change something in the UI that affects a client, uh, we'll log that action.”

Claim: The company logs any human UI change that affects a client.

“Yeah. Um, another int-interesting one. So, um, this is similar to how humans work in, in, in, uh, in our industry, is there is a process that we're tracking over ten days. And so our agents, with our kind of like long memory persistence, plus the daily reflection on what is happening, um, reflect on where the, the state of the world is at the moment.”

Claim: In Arca’s ten-day processes, agents use persistent long-term memory and daily reflection to track the current state of the world.

“And they're able to pull in the right person at that time- ... when it is-- when, when there's a signal that, that implies that that person must be pulled in.”

Claim: Arca’s agents use signals about the current state of a process to involve the appropriate person at the appropriate time.

“And all of this is, um, work that we have done to operationalize the context of how this industry works. But not just that, but how does Arca want it to work, and what are our principles of, of how work happens in this industry.”

Claim: Arca has operationalized its industry knowledge, its preferred way of working, and its principles for how work should happen in the industry.

“Um, and that is another advantage of us being sort of a, a, um, a not just a SaaS company. We're actually not a SaaS company. We are, um, we are the firm ourselves.”

Claim: Arca describes itself as the firm itself rather than as a SaaS company.

“Um, and that is another advantage of us being sort of a, a, um, a not just a SaaS company. We're actually not a SaaS company. We are, um, we are the firm ourselves. So we can standardize the way this work even happens- ... um, so that our agents can participate in standard ways that, um, that work must happen.”

Claim: Arca says that being the firm itself lets it standardize its work processes and enable its agents to participate in those workflows.

“Um, I think we are very bullish on the models from, uh, from Anthropic and, and labs like OpenAI- ... um, but less so on, on taking a dependency on a product like managed agents.”

Claim: Arca is very bullish on models from Anthropic and labs such as OpenAI, but less bullish on taking a dependency on managed-agent products.

“Um, because managed agents, one of their, their key things is how the models kind of persist and, and contribute memory- ... as to what is happening in the world, um, and all the operational knowledge that they need to be, uh, need to use to, to be successful.”

Claim: Managed-agent products can control the agents’ persistent memory of current events and the operational knowledge they need to perform successfully.

“Um, and so letting them own that part of the business is kind of like, um, a, a risk, because, um, owning that is-- Owning your intelligence is kind of, uh, gives you the ability to impact the business.”

Claim: Arca considers it risky to let an outside provider own its operational intelligence, because control of that intelligence gives the provider the ability to affect the business.

“If you, if you hand off, uh, the intelligence to, to another provider, you lose some of that control that you need to kind of like build something that is unique and tailored to your business to such a deg-great degree, um, that it can actually impact your business.”

Claim: Handing operational intelligence to an outside provider reduces a firm’s control over building a highly tailored, distinctive system that can affect its business.

“Yeah. I think, uh, I think Anthropic getting into our kind of industry is, is very validating for, for our industry. Uh, but the way I think about it is that it is very different to be an, um, a, a services firm that has access to AI tools-”

Claim: Arca views Anthropic’s entry into its industry as strong validation of that industry.

“Yeah. I think, uh, I think Anthropic getting into our kind of industry is, is very validating for, for our industry. Uh, but the way I think about it is that it is very different to be an, um, a, a services firm that has access to AI tools- ... versus a firm that was built to be AI native.”

Claim: A services firm that uses AI tools differs substantially from a firm designed from the outset to be AI-native.

“The firm with AI tools requires their users to prompt the us- the, the agents kind of to, to produce work or to do work.”

Claim: In firms that simply provide AI tools, users must prompt the agents to produce or perform work.

“Um, but here, like I said, in, at Arca, agents prompt people more than people prompt agents.”

Claim: At Arca, agents initiate requests to people more often than people initiate requests to agents.

“And it is a proactive kind of, um, service that is running for them twenty-four/seven, and it's kind of like running the business in the most standard way that we, that we expect at the level of service that we, that we, uh, that we aspire to.”

Claim: Arca’s proactive service runs continuously, 24 hours a day, and aims to operate the business according to its desired standards and service level.

“Mm. Yeah. So, um, the s- the, the way we use sandboxes is that the, the sandbox is one component of our harness. Our agent isn't actually inside of the sandbox.”

Claim: In Arca’s system, the sandbox is one component of the harness; the agent itself runs outside the sandbox.

“It uses the sandbox to offload computation.”

Claim: Arca’s agent offloads computation to the sandbox rather than running inside it.

“Um, and, uh, we've, we've built systems to make it safe for code inside of that sandbox to access our data.”

Claim: Arca has built safeguards for code running in its sandbox to access Arca’s data.

“Um, so we kind of like have this thing which translates the code it wrote into a strict format that we expect, throw away the code, and you, um, kinda instill guardrails on that format before it reaches your back end.”

Claim: Before sandbox-generated output reaches Arca’s backend, Arca converts the code into a strict expected format, discards the original code, and applies guardrails to the resulting format.

“Um, so we allow the agents kind of to use, uh, code to query our data, which means that they can like aggregate it, they can mung it, they can kind of like work with it as they, as they see fit- ... to produce such a thing like a, like a financial plan. Um, inside of the same sandbox, they produce this, you know, beautiful artifact that we, we can share with clients that has like it's a PDF of like-”

Claim: Arca’s agents can use code in the sandbox to query and manipulate company data, including to produce financial plans.

“... to produce such a thing like a, like a financial plan. Um, inside of the same sandbox, they produce this, you know, beautiful artifact that we, we can share with clients that has like it's a PDF of like- ... "Here's your financial plan, and here's how we're do- planning to kind of like approach your life, uh, financially." Um, and once that is produced is where we run our verifiers. So, um, it's a certain type of artifact that is produced from the sandbox.”

Claim: After the sandbox generates a client-shareable financial-plan PDF, Arca runs verification checks on the artifact.

“The moment that is produced, we run bunch of different, different verifications on it, um, some deterministic that check for presence of like, "Hey, did you just put like a raw account number in there? You should not do that." Um-”

Claim: Arca applies multiple checks to generated financial-plan artifacts, including deterministic checks that detect raw account numbers.

“Um, and the last thing I'll say on that is that, um, doing this Is actually, um, makes our kind of compliance, uh, better than if humans were doing it, because AI is tireless at, at, at finding even the smallest mention of a non-compliant, um, uh, piece of writing in there.”

Claim: The speaker says AI review improves Arca’s compliance compared with human review because AI tirelessly finds even small instances of non-compliant writing.

“Yeah. So what we've noticed is, um, when, when models are, um, working with financial data, right? When they're answering questions that require financial knowledge- ... they are, they're very confident in, in, uh, in the knowledge that they have. So a small example is that, uh, models will very happily use the tax brackets from twenty twenty-five- ... um, to, to come up with an analysis and be like, "All right, that is it." Um, but that is, that is obviously incorrect. And so-”

Claim: The speaker says models answering questions involving financial data can confidently use 2025 tax brackets in an analysis and present the result as settled, although doing so is incorrect.

“... um, that is one of the ways in which we're using decision models in- ... in our harness, is to, to check when the, the model is about to make a financial recommendation or, or anything, any action that requires financial rigor- ... will detect and will nudge the model to kind of use our knowledge base of verified facts about the world that is true today. And/or use their kind of web search.”

Claim: Arca’s harness uses decision models to detect impending financial recommendations or other actions requiring financial rigor and direct the model to verified current facts or web search.

“So it kind of makes it, um... This is kind of a, a, a point about the, the harness and, and owning the harness that lets you do things like this that, that help your, um, your models be, be accurate- ... where it matters.”

Claim: The speaker says owning the harness enables interventions that improve model accuracy in consequential contexts.

“No, not today. That is not a use case that we're exploring today.”

Claim: Arca is not currently exploring decision models as a way to recover when a trade fails to execute.

“So, um, so at the moment, we are not doing autonomous trading with, with agents.”

Claim: Arca’s agents do not currently trade autonomously.

“It is very much a, um, the agents will propose that we, we make trades. ... and it is fully up to kind of our, our experts to figure out if that is the right trade to be making- ... and to execute the trade.”

Claim: Arca’s agents propose trades; human experts decide whether to make each trade and execute it.

“Yeah. Um, that is another one of the advantages of us, um, being the business and not a SaaS company- ... um, is that we have, um, we have the ability to have-- to require them to give us feedback. Um, and yeah, the way that we loop the feedback back into improvements is something that you would, um, you would expect, is all of that feedback gets piped into, um, a data store-”

Claim: The speaker says Arca’s position as the business rather than a SaaS company allows it to require staff to provide feedback.

“... um, is that we have, um, we have the ability to have-- to require them to give us feedback. Um, and yeah, the way that we loop the feedback back into improvements is something that you would, um, you would expect, is all of that feedback gets piped into, um, a data store- ... that, um, that couples that feedback with the trace of the agent that produced it.”

Claim: Arca stores staff feedback alongside the trace of the agent associated with it.

“Um, and so we're able to kind of like take that, take, take that trace, um, put it, put it next to the, the feedback, and every night, we run this process where we take-- collect traces from, uh, the real world- ... from our agents in the wild, and run, run a meta agent on top that suggests changes to our harness.”

Claim: Arca nightly collects real-world agent traces and uses a meta-agent to suggest harness changes.

“So maybe tools that were mis- misleading or, um, things that were, were errored or there's missing knowledge in our knowledge base.”

Claim: Arca’s meta-agent can identify misleading tools, errors, and gaps in its knowledge base for review.

“Um, and they'll surface this information to us, and then we'll, we'll be able to kind of, uh, improve it. Um, today, we still have a, a human in, in, like a human engineer in the loop before we actually ship a prompt change.”

Claim: Arca uses surfaced trace-analysis findings to improve its system, and a human engineer reviews prompt changes before they are shipped.

“Um, I've noticed that, uh, that agents will, will make really nasty prompt changes if you leave them, leave them to it. Um- ... every little fix becomes the whole prompt. Uh, and they're, they're unable to stay meta and kind of generalize, uh, the fix into, into a prompt that generalizes.”

Claim: The speaker says agents left to modify prompts autonomously may turn each small fix into the entire prompt and fail to generalize that fix into a reusable prompt.

“Um, so but yeah, we have, we have, uh, feedback loops from our product that we collect directly as first-class product features- ... to collect evals into, into improvements.”

Claim: Arca builds product feedback collection into first-class features and uses the resulting evaluations to drive improvements.

“Yeah. So we have, um, a knowledge base of, um, of files that rep-- is represented as, as, um, as files. And, um, the way we keep that up to date is we, we've handed it to our experts to curate and update.”

Claim: Arca’s knowledge base consists of files that its experts curate and update.

“So it is a, a first-class, um, feature in our product. Uh, so on the left-hand side, there's a tab called Knowledge- ... which has the full, uh, operational context that our agents are using. Um, and as they, they work on, um, uh, work on things, uh, they're able to cite knowledge, and so it kind of like creates this, this loop of like, "Oh, we noticed that the knowledge is incorrect," um, and, um, we-- that gets surfaced to us in our trace analyses.”

Claim: Arca’s product includes a Knowledge tab containing the operational context available to its agents.

“... which has the full, uh, operational context that our agents are using. Um, and as they, they work on, um, uh, work on things, uh, they're able to cite knowledge, and so it kind of like creates this, this loop of like, "Oh, we noticed that the knowledge is incorrect," um, and, um, we-- that gets surfaced to us in our trace analyses.”

Claim: Arca’s agents can cite knowledge during their work, and indications that the knowledge is incorrect appear in trace analyses.

“And because it is a, a part of the product itself, uh, people are able to go in and, like, edit the, edit the scale, um, uh, directly.”

Claim: Arca’s product integration allows users to edit the scale directly.

“And this person was already working with an advisor. Um, and so, um, they meet with us, they tell us that, "Hey, I have this advisor already. They're charging me fifteen hundred dollars, um, uh, for, for their services."”

Claim: A prospective Arca client was already working with an advisor who charged $1,500 for their services.

“Um, "And, uh, they're not very responsive to me, so I'm, I'm chatting with you." And so, um, our agents have kind of like, um, you know, read through all of those documents. They've analyzed them. They've analyzed what the person's goals are that they state, right?”

Claim: Arca's agents reviewed a prospective client's financial documents and analyzed the client's stated goals.

“Um, "And, uh, they're not very responsive to me, so I'm, I'm chatting with you." And so, um, our agents have kind of like, um, you know, read through all of those documents.”

Claim: A prospective Arca client said their existing financial advisor was not very responsive.

“Um, they've also analyzed a statement of, like, who this current advisor is, uh, right, and how much they're, they're charging them. And, um, they discover that, that there is, there are fees that this person was charging them that they weren't aware of, uh, and surfaces it, it to them, um, and kind of like makes them not only aware of, um, of the fact that they're being overcharged, but also aware that that person wasn't even fiduciary.”

Claim: Arca's agents analyzed information about a prospective client's existing financial advisor and the advisor's charges.

“Um, they've also analyzed a statement of, like, who this current advisor is, uh, right, and how much they're, they're charging them. And, um, they discover that, that there is, there are fees that this person was charging them that they weren't aware of, uh, and surfaces it, it to them, um, and kind of like makes them not only aware of, um, of the fact that they're being overcharged, but also aware that that person wasn't even fiduciary.”

Claim: Arca's agents identified previously undisclosed fees charged to a prospective client and surfaced them to the client.

“Um, they've also analyzed a statement of, like, who this current advisor is, uh, right, and how much they're, they're charging them. And, um, they discover that, that there is, there are fees that this person was charging them that they weren't aware of, uh, and surfaces it, it to them, um, and kind of like makes them not only aware of, um, of the fact that they're being overcharged, but also aware that that person wasn't even fiduciary. Um, so they're not fu-- uh, obligated to act in their interests.”

Claim: Arca's agents informed a prospective client that the client was being overcharged and that the existing advisor was not a fiduciary.

“Um, so they're not fu-- uh, obligated to act in their interests.”

Claim: A financial advisor who is not a fiduciary is not obligated to act in the client's interests.

“So it is-- This is the sort of, like, attention to detail that our agents bring about in, in-- when they're analyzing a financial picture-”

Claim: Arca's agents bring attention to detail when analyzing a person's financial picture.

“... um, that, that sort of, um, would not have been pos-possible if we didn't have, um, have these. There's several examples of, um, proactivity, uh, that our agents kind of like produce of, like, someone has a deadline coming up. Or they're at risk of missing their RMD or, um, that kind of thing. And so, um, this is difficult at scale for humans to keep track of, but it is the, the default behavior of, uh, of our agents at, at Arca.”

Claim: Arca's agents proactively identify financial situations such as upcoming deadlines and risks of missing required minimum distributions (RMDs).

“... um, that, that sort of, um, would not have been pos-possible if we didn't have, um, have these. There's several examples of, um, proactivity, uh, that our agents kind of like produce of, like, someone has a deadline coming up.”

Claim: The speaker said Arca's attention to detail in financial analysis would not have been possible without its agents.

“Or they're at risk of missing their RMD or, um, that kind of thing. And so, um, this is difficult at scale for humans to keep track of, but it is the, the default behavior of, uh, of our agents at, at Arca.”

Claim: The speaker said tracking financial deadlines and risks is difficult for humans at scale, whereas proactively tracking them is the default behavior of Arca's agents.