Proving Alpha: How a Hedge Fund Measures AI's ROI - Hari Kumar, Contour Asset Management
Slide chapters
7 slide and demo groups · Timestamps are approximate.
Read the talk
Proving Alpha: How a Hedge Fund Measures AI’s ROI
Hari Kumar explains how Contour Asset Management compares research workflows, tracks cost per useful insight and counts the time lost to broken tools while waiting for investment outcomes to emerge.
From a talk by Hari Kumar
At a glance
Ideas worth remembering
Document research with and without AI at the level of data sources, delegated activities, conclusions and time spent. Contour’s concrete example follows one company between two earnings reports.
Process improvements can become visible before investment returns. A thesis may take six months or a year to affect a stock price.
Agents can preserve the usefulness of earlier research by checking shelved ideas against the market and bringing them back to the analyst, portfolio manager and trader.
Cost per useful insight connects analyst effort, the tools used, the ideas produced and their eventual outcomes. Tool failures and learning time belong in that accounting.
The spending happens before the investment result
An AI tool can change an analyst’s work today while the investment thesis takes months to affect a stock price. That delay makes a straightforward question surprisingly difficult: what is the fund getting for its spending? Hari Kumar, head of AI at Contour Asset Management, describes the problem at a fundamental long/short hedge fund with roughly $6 billion under management and fifteen to twenty people, focused largely on technology, media and telecom.
Kumar’s mandate starts with research and extends into trading, operations, compliance and finance. The measurement problem has two parts: whether AI helps produce better ideas and better working processes, and whether those changes eventually contribute to profit and loss, or P&L. The first can become visible while the second is still developing. That distinction drives Contour’s experiments.
Suggest correction
This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.
Compare the research process over an earnings cycle
Contour has around ten analysts, each covering a portion of its investment universe. Kumar describes the small-fund pattern of one analyst following thirty to forty companies, with established habits for developing and testing investment theses. Introducing AI raises a counterfactual question: would the analyst have reached the same conclusion without it? Contour approaches that question through what Kumar calls live A/B tests, documenting research with and without the new tools. 3:05
One comparison follows a single company over a three-month period, from one earnings report to the next. The baseline records the analyst’s workflow without LLMs and without additional workflows such as agent-conducted consumer surveys or tools for expert interviews. The record includes which data sources the analyst finds and what conclusions follow. Contour then runs the research process with AI tools, involving the portfolio manager, or PM, and introduces tools individually as they become available.
What changes when the agent takes on another part of an expert interview? Kumar separates two arrangements:
- Agent interview, human synthesis: An agent conducts the expert call, and the analyst synthesizes the notes.
- Agent interview and synthesis: The agent also synthesizes the notes, leaving the analyst to draw the final conclusions.
The distinction makes the division of work inspectable. Conducting a call, condensing its contents and forming an investment judgment are separate activities; automating one does not establish the value of automating the next.
The practical comparison asks how long it would have taken to source the information manually and whether the resulting conclusions change. Kumar describes the approach as useful, but gives no numerical treatment effect or details of random assignment. These are documented workflow comparisons; they do not establish a measured causal improvement in investment returns.
Suggest correction
This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.
Measure useful changes before the thesis pays off
The earnings cycle gives the comparison a workable observation window, but a thesis may need six months or a year to show up in a stock price. Contour therefore looks first at the work itself: whether consumer surveys improve, whether investment models become easier to update and whether information takes less time to obtain. Those changes can be evaluated before the market resolves the thesis.
The uncomfortable alternative is that another tool simply adds “more thrash” to the analyst’s day. A workflow can generate more material while making research harder to complete. Contour’s question is whether the tool changes how someone works usefully enough to justify the extra activity.
Suggest correction
This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.
Keep a shelved idea from missing its moment
Fundamental research starts with a thesis, but it also requires letting new evidence loosen the analyst’s attachment to that thesis. Contour uses idea-generation agents to add further considerations to ideas organized by company and analyst. An idea may begin with alternative-data observations or an industry expert’s account; the additional work examines how it might develop over the next few months, or why the market might take longer to reflect it.
A different failure happens after an idea has already received attention. An analyst works on it for a day or two, the PM finds something else more pressing, and the idea disappears into the notes. Months later, the team may realize that it missed an attractive price window. The research was preserved, but the opportunity to use it was lost.
The proposed remedy is to have agents track those notes against the market and resurface an idea when it becomes interesting again. The alert can reach the original analyst, the PM and the trader. How does a forgotten note become shared attention? The diagram traces the two inputs—past research and market conditions—into that reassessment, then shows the separate recipients. The output is a reason to revisit the work; Kumar does not describe automatic trade execution.
This expands the job beyond generating fresh ideas. A fund must choose among more possible investigations than its analysts can pursue, and a previously uninteresting idea can become useful at a different price or time. Kumar describes entry and exit timing as an area Contour is actively exploring with agents, including positions that are no longer top of mind. 8:18
Resurfacing research when the market changes
Read the diagram as text
- Saved analyst notes. inputs — Includes ideas investigated briefly and then set aside.
- Market conditions. inputs — The context against which an old idea is reconsidered.
- Agent reassessment. Tracks the notes against the market and asks whether the idea is interesting now.
- Analyst. recipients — Receives a prompt to revisit prior research.
- Portfolio manager. recipients — Receives the resurfaced idea.
- Trader. recipients — Receives the resurfaced idea.
- Saved analyst notes → Agent reassessment: Supplies earlier research.
- Market conditions → Agent reassessment: Supplies market context.
- Agent reassessment → Analyst: Flags an interesting idea.
- Agent reassessment → Portfolio manager: Flags an interesting idea.
- Agent reassessment → Trader: Flags an interesting idea.
Suggest correction
This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.
Bring expert accounts and spending data together
Funds already buy specialized datasets, including satellite mapping and granular credit-card data. The integration problem is that each can remain isolated from the rest of the research. Kumar’s concrete example is credit-card panel data supplied by Yipit: Contour feeds it into agents through MCP and API access alongside expert-call transcripts and research notes.
Combining those inputs lets the research consider how an expert’s account interacts with observed spending data. The change is from separately consulting a transcript and a dataset to bringing both into the agent’s analysis. Kumar describes that integration as implemented at Contour, without specifying a matching algorithm or a particular investment conclusion produced by it.
Suggest correction
This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.
Track cost per useful insight
Subscription and AI spending capture only part of the research cost. Contour also tracks what analysts spend their days doing: time spent on each company, insights generated per day and the eventual P&L outcomes of decisions. A fundamental fund may make a few trades a day while expecting those investments to develop over months, so keeping the research history connected to later outcomes matters. 10:35
Contour calls one of its metrics “cost per useful insight.” Its supporting questions connect effort to consequence:
- Time: How much analyst time went into researching the company and generating the idea?
- Usefulness: Where did the idea eventually lead, and which ideas proved interesting?
- Tools: Which tools helped produce it?
Kumar does not specify a formula, a rule for classifying usefulness or a measured value. The practical contribution is the connected record: daily research effort can later be examined alongside the ideas it produced and their outcomes.
Suggest correction
This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.
Give trading and operations more time for judgment
The research experiments sit within a wider effort to improve how the firm works. Trading, operations and finance can free time for better processes and contribute to the fund’s understanding of the market. Kumar describes two distinct directions:
- Trading: Contour measures entry and exit timing and examines how positions are spread across industries and individual companies. Agents track those companies to keep traders informed. The aim is for traders to contribute ideas to analysts and associates as well as execute orders.
- Operations and compliance: Work is distributed across Bloomberg, order-keeping systems and other provider tools. Contour is starting to integrate activities such as order reconciliation and vendor-information tracking into one tool, reducing the need to visit multiple platforms for different pieces of the order flow.
Individual providers may already offer helpful AI features, while the work of joining them remains difficult. Consolidation is a priority because repetitive handling consumes time that people in trading and operations could apply to generating ideas, understanding the market and judging timing. Kumar presents this as a direction for reallocating skills, rather than a quantified return from a completed operations rollout.
Suggest correction
This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.
Make existing subscriptions fit each analyst’s work
The next question is how to get more value from data the fund already buys. Kumar names specialist subscriptions such as AlphaSense and Yipit alongside broader platforms such as Rogo and Hebbia. In his AlphaSense example, an analyst traditionally visits the provider’s portal, retrieves information and uses it in an existing investment model. He describes Rogo as taking integration further by bringing data into the model and providing agents the analyst can chat with.
Contour’s in-house ambition keeps the underlying subscriptions and brings their data into an internal platform tailored to each analyst’s habits and model patterns. MCP and API connections provide access; the intended additional value comes from fitting that access to how a particular analyst works. The platform is meant to learn over time from data use and investment-model updates, though Kumar does not describe the learning mechanism. This is a customization goal, not a demonstrated model-training result.
Suggest correction
This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.
Count the time spent fixing and understanding the tools
An internal platform also brings an operating cost that is easy to leave out of the ROI calculation. Model calls fail. A connector can break every couple of days. Research pauses while someone fixes the problem, and some bugs require help beyond what the technical team can resolve alone. Contour counts those interruptions and the time taken away from other work. 16:40
Learning how the tools work also takes analyst time. That cost belongs beside the time used to generate a valuable insight: a faster research step can still carry setup, explanation and repair work elsewhere. Kumar says the time spent understanding the tools is decreasing as Contour proceeds, but it remains something the firm measures.
Kumar closes with a bet on a unified layer serving multiple analysts and tasks, especially at smaller fundamental funds. Its value depends on making the firm’s existing data and people more effective while accounting for the effort required to keep that layer useful. The maintenance burden is part of the investment.
Suggest correction
This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.
Resources
Related talks
- Containment Is Lying to You: Measuring AI Agents Like Employees — Druid AI
Develops the measurement question further by comparing agents against human baselines and distinguishing activity metrics from business value.
- How to Know When an AI Agent Can Work Unsupervised — Evan Kormos, Coinbase
Explores how to evaluate bounded agent tasks in regulated workflows, including process metrics, traces, human review and exception handling.
Read the complete timestamped transcript
7 slide groups. Extracted from the recording; expand builds to follow changes. Timestamps are approximate.
Measuring the impact of AI in fundamental investing
Hari Kumar introduces Contour’s effort to measure AI’s return, starting with investment research: whether the firm gains useful ideas and operational improvements, alongside financial results.
Text in image
Measuring the impact of AI in fundamental investing CONTOUR / AI & INVESTING OCTOBER 2026 What we are trying to capture, and where are are heading. HARI KUMAR CONTOUR ASSET MANAGEMENT 01 Interface text: AI Engineer NEW YORK PRESENTED BY arize OCTOBER 13, 2026 BREAKOUT SESSION Turing Room AI Engineer NEW YORK CC
- 0:12
My name's Hari. Um, I am the head of AI at Contour Asset Management, which is a hedge fund based in the city. It's been around for a little over a decade now, um, but I've been here for the past year-ish. Um, we're a fundamental long/short shop, um, around like fifteen, twenty people, six billion AUM um, and mostly invest in the TMT space. Um, so a lot of our investors are pretty tech-forward by nature, but want to bring a little bit
Measuring the impact of AI in fundamental investing
Hari Kumar introduces Contour’s effort to measure AI’s return, starting with investment research: whether the firm gains useful ideas and operational improvements, alongside financial results.
Text in image
Measuring the impact of AI in fundamental investing CONTOUR / AI & INVESTING OCTOBER 2026 What we are trying to capture, and where are are heading. HARI KUMAR CONTOUR ASSET MANAGEMENT 01 Interface text: AI Engineer NEW YORK PRESENTED BY arize OCTOBER 13, 2026 BREAKOUT SESSION Turing Room AI Engineer NEW YORK CC
- 0:42
more, um, AI workflow into the investment firm itself and see, okay, we've had all these processes done for many years in a very similar way. How can we kind of shake things up and use AI to our advantage, especially when it comes to the research process, but then also bring that into trading, operations, compliance, finance, pretty much all parts of the firm. So came in in the past year, and basically, my mandate was, how can I show
- 1:12
impact across all areas of the firm, largely starting with this research process, right? And as probably all of you guys in this room can understand and appreciate, it's a really big challenge to show, hey, we're spending this amount of money and time on AI spend and impact. What are we actually gaining from it? Both from a P&L standpoint, which is to come, um, but also just from a more operational and
- 1:42
are we getting good ideas perspective too. The ROI capture from our AI investment is a really big challenge to understand and capture, but I wanna talk a little bit about some of the ways at-- in the investing landscape we're able to start to think about, hey, we've put in this amount of money into our AI spend. What are we getting out of it?
- 2:06
Um, cool. So a few things that we are thinking about here, and intentionally trying to keep this a little bit vague and more in terms of the operational and process parts of it, but down to dive into any technical stuff either at the end of the session or, um, afterwards as well. So what we're trying to do is basically bring out live A/B tests for some of the research workflows that our analysts have. So we have around ten analysts that each cover
What would have happened without AI?
Contour compares documented research workflows with and without AI, tracking conclusions and sourcing time to distinguish better judgment from faster execution. A clean counterfactual remains difficult.
Text in image
What would have happened without AI? Actually mapping out A/B Tests CONTOUR / AI & INVESTING THE MEASUREMENT PROBLEM THE FUNDAMENTAL QUESTION 01 INVESTMENT DECISIONS Would we have found the same idea or reached the same conclusion? 02 RESEARCH ACCESS How long would sourcing the information have taken manually? 03 OUTCOMES Is the benefit better judgment, faster execution, or both? No clean control group. No easy counterfactual. 02 OCTOBER 13, 2026 BREAKOUT SESSION Turing Room Interface text: CC
- 2:35
different parts of our coverage, right? Which is how it works at a lot of these, um, hedge funds, especially on the smaller side, where you have maybe one person that covers thirty to forty names. And they've had their, uh, existing processes in terms of how they think about their investment thesis, how they actually come up with new ideas, and then validate and test out over the long term whether those ideas actually came to fruition. But now in a AI and more agentic world when it comes to
- 3:05
investing, the challenge is would you have come to the same type of conclusion with the AI tools or without the AI tools as well? And a process that we've done in the past few months that I think has been pretty successful, has been actually creating live A/B tests by recording and documenting their entire process that they would have done in a pre-AI world and how they're doing it as we roll out more workflows and tools in the company itself.
What would have happened without AI?
Contour compares documented research workflows with and without AI, tracking conclusions and sourcing time to distinguish better judgment from faster execution. A clean counterfactual remains difficult.
Text in image
What would have happened without AI? Actually mapping out A/B Tests CONTOUR / AI & INVESTING THE MEASUREMENT PROBLEM THE FUNDAMENTAL QUESTION 01 INVESTMENT DECISIONS Would we have found the same idea or reached the same conclusion? 02 RESEARCH ACCESS How long would sourcing the information have taken manually? 03 OUTCOMES Is the benefit better judgment, faster execution, or both? No clean control group. No easy counterfactual. 02 OCTOBER 13, 2026 BREAKOUT SESSION Turing Room Interface text: CC
- 3:35
So for example, what we did is for one specific name when it came to a three-month cycle, so from one earnings report to the following one, we tracked if they didn't use any AI tools. And this is at first just, "Hey, if you're not using any LLMs , what does your workflow look like?" But then as we've introduced more, um, tools at the hedge fund itself, like, okay, you have some tools that help you do expert call interviews better or
- 4:05
ones that help you field a consumer survey that's done by an agent instead of a human. What are your takeaways without things like that? And then we're documenting, okay, you've done this process without AI, but also without these additional tech workflows that we're building out. What are the data sources that you're finding, and what are the conclusions you're coming out with? Then we run that same process involving our PM , including some of these AI tools, and we're building them one by one with
- 4:35
different things that we're coming out with, right? So for example, when it comes to expert call interviews, one thing that's been pretty helpful is you have expert calls that are run with an agent, and then you synthesize the notes yourself. How does that process work when you do it with an agent work-- versus if you have the agent synthesize the notes for you, and then you come up with your conclusions at the very end of it, right? So it's kind of thinking about the question of how long would it have taken to source some of
- 5:05
this information manually versus if you were able to do it in a more agentic manner. This, this is definitely a big challenge when it comes to the investing landscape because a lot of the work that we're doing is just based off of idea generation. So you're coming up with new ideas and new theses, and you wanna see, okay, can they be proven out in a short-term scale, which is a little less of what we do, but at least in like the three-month window. Or, or are they things that take six
- 5:35
months or a year to manifest in terms of an actual, um, impact to stock price, right? So the challenge with something like this is you can't show ROI immediately. It's more so from a process-based manner, right? So are we improving the process of how we're conducting consumer surveys? Are we improving the process of how we're updating our models? And one by one, we're trying to bring out a little bit more of, um-
- 6:06
uh, ROI measurement when it comes to, okay, are we actually changing the way someone's working or are we just adding more thrash to their process?
Better thinking, not just faster work
AI helps analysts challenge an investment thesis by combining expert notes, surveys, and alternative data. Measure the added insight through novel ideas, differentiated sources, and ideas acted upon.
Text in image
CONTOUR / AI & INVESTING 01 / RESEARCH WORKFLOWS MEASURE THE INCREMENTAL INSIGHT Better thinking, not just faster work. How to Better Challenge a Thesis Goal here is to surface sources an analyst would not otherwise use. Use Cases: - Expert Calls and Consumer Survey Notes - Idea Generation - Alternative Data → Second/Third-Order Connections THE QUESTION TO ASK “What did you consider that you wouldn’t have before?” Track novel ideas, differentiated sources, and ideas acted upon. The value is additional signal: not simply hours saved. 03 OCTOBER 13, 2026 BREAKOUT SESSION Turing Room Interface text: CC
- 6:19
So talked a little bit about these use cases, but maybe just to touch a little bit more on them. The main goal when it comes to, um, fundamental investing is thinking of a thesis, but also as I've learned, it's to not be so anchored to a thesis that you come up with initially and actually trusting some of the data, but also the expert and other insights that you're getting to remove some of those anchors as well, right? And this is where I think that there's a lot of impact measurement
- 6:49
when it comes to AI tooling that can be built. So example, um, we have manual idea generation agents that are being run to take different ideas that are being come up with per name, per analyst, and adding more layers in terms of, okay, you've come up with a few different data points from different alternative datasets, or you've talked to an expert in the industry to formulate some of that idea. Let's layer on a few more
Better thinking, not just faster work
AI helps analysts challenge an investment thesis by combining expert notes, surveys, and alternative data. Measure the added insight through novel ideas, differentiated sources, and ideas acted upon.
Text in image
CONTOUR / AI & INVESTING 01 / RESEARCH WORKFLOWS MEASURE THE INCREMENTAL INSIGHT Better thinking, not just faster work. How to Better Challenge a Thesis Goal here is to surface sources an analyst would not otherwise use. Use Cases: - Expert Calls and Consumer Survey Notes - Idea Generation - Alternative Data → Second/Third-Order Connections THE QUESTION TO ASK “What did you consider that you wouldn’t have before?” Track novel ideas, differentiated sources, and ideas acted upon. The value is additional signal: not simply hours saved. 03 OCTOBER 13, 2026 BREAKOUT SESSION Turing Room Interface text: CC
- 7:18
conclusions or parts of the process to think about when it comes to how are you actually, how are you actually understanding what that idea is going to materialize in the next few months, or if it's something that it takes even longer to take, uh, to take into account into the market. Additionally, we're thinking a little bit more as to one of the biggest problems that we're seeing is you take an idea or you have an idea, you might work on it for a couple of days, and then that idea might not be
- 7:48
interesting to the PM or something else takes precedence, and you lose that idea, and it doesn't come back for months, and then maybe you miss like a key price window. So having agents that are also tracking through all of the notes that you've been taking, even if you only worked on an idea for a day, tracking it against the market and trying to see, okay, is this now an idea that's interesting, and should we flash that as a point to not just you, but the PM and the trader as well? And bringing out more quick work when it comes
- 8:18
to you've had a bunch of different things you've worked on, and maybe it's been weeks since you've looked at it. Now let's get back to it because it's interesting again, right? I think that's a big problem that a lot of hedge funds have had, where you have a bunch of different things that you can be working on. You have to prioritize, but the timing is actually the most important thing as to when you wanna strike when it comes to an interesting point in the market, right? So building agents that are not only doing idea generation, but thinking about timing
- 8:48
and when you should enter or exit a position, even if it's not something that has been top of mind for you, is something that we've been really thinking about as well. Um, and I guess the last point that I wanna make here, thinking a little bit more about the alternative data options when it comes to investing workflows, right? A lot of funds spend quite a good sum of money when it comes to acquiring really new and novel datasets, right? And I think in this age, there's a bunch of different providers that are coming out with very nuanced
- 9:18
solutions, even when it comes to things like satellite mapping or really granular credit card data. Those things are helpful, but they're only helpful when they're all brought together, right? So if you think about something like a Yipit, which provides credit card panel data, that lives in isolation. But now with MCP and API access for things like Yipit, you can feed into our agents, which is kind of what we've done. So it takes into account things like your expert call or notes transcripts, but also
- 9:48
layers in different datasets when it comes to credit card panels or spend data to see how those two things interact. Where before those would have been differentiated, now we're trying to curb that issue as well.
Same data spend. Different utility.
MCP and APIs turn subscriptions into agent calls that build reusable research context. Contour tracks analyst time and cost per useful insight while investment outcomes take months to materialize.
Text in image
CONTOUR / AI & INVESTING 02 / DATA ACCESS THE ECONOMICS OF SUBSCRIPTIONS Same data spend. Different utility. BEFORE Subscription → manual search → extraction → research WITH MCP / APIS Subscription → agent calls → reusable research context THE NEW VENDOR MATH Utilization / calls / citations / adoption / cost per useful insight Programmable access becomes a competitive requirement. 04 Interface text: AI Engineer
- 10:05
Additionally, I think a challenge with measuring ROI is the time aspect of it. So you've spent a lot of money when it comes to acquiring some of this tooling, when it comes to research, model updates, AI spend. But there's a lot of time when it comes to what is an analyst spending their day on, which is the main thing that we're thinking about, right? So what we've also done here, which I think has been pretty useful, is develop metrics when it comes
Same data spend. Different utility.
MCP and APIs turn subscriptions into agent calls that build reusable research context. Contour tracks analyst time and cost per useful insight while investment outcomes take months to materialize.
Text in image
CONTOUR / AI & INVESTING 02 / DATA ACCESS THE ECONOMICS OF SUBSCRIPTIONS Same data spend. Different utility. BEFORE Subscription → manual search → extraction → research WITH MCP / APIS Subscription → agent calls → reusable research context THE NEW VENDOR MATH Utilization / calls / citations / adoption / cost per useful insight Programmable access becomes a competitive requirement. 04 Interface text: OCTOBER 13, 2026 BREAKOUT SESSION Turing Room
- 10:35
to, hey, these are the insights that have been generated by an analyst per day. What was the time spent on that? So tracking time spent on each name, and what were the eventual P&L outcomes from some-- from a decision like that. That is something that is good in investing landscape because you have a very, very obvious success metric when it comes to P&L. But it takes a lot longer than other industries to actually materialize,
- 11:04
especially from a more fundamental shop where you're not trading by the day, but you're making a few trades a day that are expected to materialize over months-long horizons, right? So understanding where analysts' time goes is really valuable. And tracking things like a cost per useful insight, which is a metric that we've come up with, really helps us understand, okay, you've spent your time to generate these ideas, and those are all being tracked per day. Where do those
- 11:34
ideas eventually land? What were the interesting ones? And what were the tools that were taken to come up with an idea like that?
Bring the same discipline to the rest of the firm.
Contour extends its research measurement discipline to trading, operations, and finance. Repeatable workflows enable testing against actual history; consolidating fragmented tools aims to free staff for market insight and timing.
Text in image
Bring the same discipline to the rest of the firm. CONTOUR / AI & INVESTING 03 / FIRMWIDE IMPACT BEYOND THE RESEARCH DESK 01 Trading Order history, execution timing, and historical lookbacks 02 Operations Repetitive work, exceptions, and process reliability 03 Finance / Analytics Query, reconcile, and understand internal history As workflows become repeatable, more outcomes can be tested against actual history. 05 Interface text: OCTOBER 13, 2026 BREAKOUT SESSION Turing Room
- 11:44
So that largely touches on the research side of workflows when it comes to measuring, um, our ROI from not just AI spend, but general tech investment. But we want to also improve this across the entire firm, right? It's really important to make sure that we do quality research. That is how you get ahead. But if you have better and more involved trading, as well as stronger ops and finance, that just helps the firm at large to free up time to improve processes, right?
- 12:15
So for example, on the trading side of things, we're doing a lot more when it comes to measuring our entry and exit times, thinking about how our position is spread across multiple different industries and multiple names within an industry, and having AI agents track those names to make sure that our traders are on top of it and understand, okay, these are interesting insights. So bringing them much more into the analyst world because their time should not just be spent on executing orders. It should also be
- 12:45
spent on bringing idea generation, feeding ideas to our analysts and associates as well. Additionally, on the ops side, this is something that I think a lot of, um, general companies can appreciate, but it's been pretty interesting to see it from a hedge fund landscape. Many of the workflows that are being done are basically set within Bloomberg or set within your, um, order keeping agent. And those, while they've come up with some cool AI tools
- 13:15
to help improve the process, it's really challenging to integrate those together, which is something that we're starting to do here. So how can you take all parts of the ops and compliance workflow and marry them into one, one specific tool that our ops and compliance, um, functions can use so they're not just reliant on external providers to give them piecemeal solutions to, hey, this is how you can do order reconciliation, or this is how you can, uh,
- 13:45
track vendor info. The goal that we're trying to create here is obviously minimize repetitive work. That's what everyone here is thinking about. But specifically from financial operations, that is something that can really be consolidated into one platform, where right now you have to kind of go into many different platforms for different pieces of the order flow. So bringing those all into one is, um, is a pretty big priority for us. And I think not necessarily to the point of this talk, a way to
Bring the same discipline to the rest of the firm.
Contour extends its research measurement discipline to trading, operations, and finance. Repeatable workflows enable testing against actual history; consolidating fragmented tools aims to free staff for market insight and timing.
Text in image
Bring the same discipline to the rest of the firm. CONTOUR / AI & INVESTING 03 / FIRMWIDE IMPACT BEYOND THE RESEARCH DESK 01 Trading Order history, execution timing, and historical lookbacks 02 Operations Repetitive work, exceptions, and process reliability 03 Finance / Analytics Query, reconcile, and understand internal history As workflows become repeatable, more outcomes can be tested against actual history. 05 Interface text: OCTOBER 13, 2026 BREAKOUT SESSION Turing Room
- 14:16
really capture ROI, but to understand how can you take people who are working in trading and ops right now and use their skills to improve the main goal of the fund, which should be to create, create new ideas, understand the market, and focus on timing.
From chat tool to operating layer.
Contour aims to bring subscribed data into an internal platform tailored to analysts’ habits, turning standalone chat and manual retrieval into embedded agents and repeatable, auditable workflows.
Text in image
From chat tool to operating layer. CONTOUR / AI & INVESTING WHERE THIS IS GOING THE DIRECTION OF TRAVEL Standalone chat → Agents embedded in workflows Manual retrieval → Programmable data access One-off answers → Repeatable, auditable processes The differentiator: data access, workflow design, and human judgment. 06 Interface text: AI Engineer NEW YORK PRESENTED BY arize AI Engineer NEW YORK OCTOBER 13, 2026 BREAKOUT SESSION Turing Room
- 14:35
Okay. The, the last piece that I wanna talk about here is thinking a little bit about the existing spend that we have, um, when it comes to different data sources that we're working with, and how we can kind of up-level that when it comes to hedge fund workflows. So right now, a lot of hedge funds will spend on-- either they're spending on different piecemeal solutions like an AlphaSense for, for data, a Yipit for alt data, or they're spending with, um,
From chat tool to operating layer.
Contour aims to bring subscribed data into an internal platform tailored to analysts’ habits, turning standalone chat and manual retrieval into embedded agents and repeatable, auditable workflows.
Text in image
From chat tool to operating layer. CONTOUR / AI & INVESTING WHERE THIS IS GOING THE DIRECTION OF TRAVEL Standalone chat → Agents embedded in workflows Manual retrieval → Programmable data access One-off answers → Repeatable, auditable processes The differentiator: data access, workflow design, and human judgment. 06 Interface text: AI Engineer NEW YORK PRESENTED BY arize AI Engineer NEW YORK OCTOBER 13, 2026 BREAKOUT SESSION Turing Room
- 15:06
uh, holistic solutions that use, um, third party. So something like buying a Rogo membership or investing in like Hebbia, right? Those are really great solutions, and I think the, the goal of this eventually that we're trying to build in-house is how can you kind of bring all these tools that different analysts use at a hedge fund into one platform, but also up-level the spend that we have right now. So for example, with an
From chat tool to operating layer.
Contour aims to bring subscribed data into an internal platform tailored to analysts’ habits, turning standalone chat and manual retrieval into embedded agents and repeatable, auditable workflows.
Text in image
From chat tool to operating layer. CONTOUR / AI & INVESTING WHERE THIS IS GOING THE DIRECTION OF TRAVEL Standalone chat → Agents embedded in workflows Manual retrieval → Programmable data access One-off answers → Repeatable, auditable processes The differentiator: data access, workflow design, and human judgment. 06 Interface text: AI Engineer NEW YORK PRESENTED BY arize AI Engineer NEW YORK OCTOBER 13, 2026 BREAKOUT SESSION Turing Room
- 15:36
existing, let's say, AlphaSense subscription, which is a big, um, model and call plus data provider. Before, a lot of analysts will just use the AlphaSense portal and pull from there and work within their existing models. Things like Rogo will kind of take that one step further by allowing you to have all that data within your model itself, and you can have a platform that you chat with Rogo agents for. What we're trying to do is still use
- 16:06
things like an AlphaSense or Yipit tool, but bring them into our own platform internally that learns from our analysts u-- uh, analyst habits, and is programmed specifically for each individual analyst and their model pattern, so it's not all as one. And how can you increase the value of a subscription? A lot of these companies obviously are having like MCPs and APIs that are being pulled out, but the goal should be to learn continuously from all the different types of data and the
- 16:36
model updates that we're getting from this as well.
Measuring ROI based on a variety of factors
AI’s investment gains must cover time spent fixing failures and learning tools. Contour weighs better signals and research capacity against these costs, measuring quality, outcomes, and failure rates alongside usage.
Text in image
Measuring ROI based on a variety of factors CONTOUR / AI & INVESTING A WORK IN PROGRESS AN HONEST ACCOUNTING THE UPSIDE New signal. Better decisions. Expanded research capacity. THE REAL COSTS Model errors. Broken connectors. Verification. Rework. Measure usage alongside quality, outcomes, and failure rates. Additionally explore longer-term tracking of investment ideas. 07 Interface text: CC OCTOBER 13, 2026 BREAKOUT SESSION Turing Room
- 16:40
And then to wrap up here, I think the, the last point of the ROI equation too has been a pretty interesting one on my end when it comes to, okay, you're bringing in new tooling to people who are very tech-forward for investors, but you understand that there are going to be places where this break down, where model calls break, or the connector breaks down every couple days, and then it's your job to fix it, right? So part of the ROI equation that we've been taking into account at Contour is trying to
- 17:10
understand, okay, there are these pauses that we're getting, but also what is the time that is being taken away trying to fix bugs that someone from the tech side cannot fix itself, right? So that's something that I think is pretty important to consider when it comes to, hey, we're involving AI in investing workflows. It's all really exciting and people are very into the concept of it. But we also need to understand, okay, when it comes to time spent for, let's say, valuable insight, what is the time detracted to understand how these
Measuring ROI based on a variety of factors
AI’s investment gains must cover time spent fixing failures and learning tools. Contour weighs better signals and research capacity against these costs, measuring quality, outcomes, and failure rates alongside usage.
Text in image
Measuring ROI based on a variety of factors CONTOUR / AI & INVESTING A WORK IN PROGRESS AN HONEST ACCOUNTING THE UPSIDE New signal. Better decisions. Expanded research capacity. THE REAL COSTS Model errors. Broken connectors. Verification. Rework. Measure usage alongside quality, outcomes, and failure rates. Additionally explore longer-term tracking of investment ideas. 07 Interface text: CC OCTOBER 13, 2026 BREAKOUT SESSION Turing Room
- 17:40
tools work as well? And that's something where that time is decreasing as we go forward, but it's a metric that we're still considering. Um, so I think to wrap up, in investing workflows, measuring ROI is definitely a huge challenge for us, but it's really important to be involved in this space because as AI adoption increases, especially at a lot of these smaller fundamental shops, having one unified layer that's doing a lot of different tasks across analysts is gonna be really, really
- 18:10
valuable. Um, I think that's all the time I have, but if anyone has any questions after the talk, would love to connect. Thank you.