Acuity Knowledge Partners began as a tech shop in 2005. Initially focused on front-office applications with a modest team and a limited range of solutions, Acuity Knowledge Partners experienced a significant transformation around 2011-2012. This period marked a shift driven by major industry changes, including the implementation of regulations like MiFID II in Europe, which impacted research, sales, and trading, and increased demands from the buy side for more frequent and detailed reporting. The simultaneous decline in computing and data costs further accelerated this transformation.
From 2011 onward, Acuity Knowledge Partners embraced these changes, transitioning from on-premises solutions to cloud-based systems and evolving from traditional machine learning to cutting-edge AI technologies. This evolution was fueled by a growing need to develop tech and data talent to meet the expanding demands of the financial services sector. Despite these advancements, the company has remained true to its core principle established in 2005: delivering solutions that are contextually relevant to banks and asset managers. Acuity’s commitment to providing innovative, tailored solutions continues to drive their success, emphasizing the importance of achieving a tangible return on investment in the rapidly evolving financial landscape.
Recently, Deepak Stephen, Head of Data Science at Acuity Knowledge Partners spoke to Tech Achieve Media to share more insights into what lies ahead for the firm in this ever-evolving tech era.
TAM: When it comes to financial institutions, what are some common problems they face in relying on data for actionable insights? And how do you help them address these issues?
Deepak Stephen: Our area of expertise is in the front office part of financial markets. This includes investment banks, some functions of commercial banks, asset managers, and private markets. Over the last decade, these institutions have faced a deluge of data. For example, industry studies, such as those from Seagate and IDC, estimate that by next year, around 175 zettabytes of data will be generated.
These organizations were originally designed for legacy operations that are regulated and audited, making this a significant paradigm shift for them. Unlike a B2C environment, changes in these institutions cannot happen drastically; they require time, effort, and a sustainable approach. CXOs, including chief technology officers, chief data officers, and chief analytics officers, need to stay competitive while acting responsibly.
The deluge of data presents a problem. Different organizations have different strategies to handle it. Some use in-house teams to maintain data lineage, but in developed markets, talent becomes an issue due to population caps and a shortage of skilled professionals. These organizations seek stable solutions that can guide them through this journey. Some opt for partnering with platforms or CRM systems instead of building from scratch, while others need people on-site for audited processes.
When classifying data, there are traditional structured, numeric data sets and newer types, such as textual data, used in investment processes and vetting. Managing this data requires people, reliable systems, and auditability of everything, including code and data. This is where we differentiate ourselves from competitors; we understand these processes and the unique needs of different financial institutions.
For example, the way a hedge fund resolves data challenges differs from that of a commercial bank. However, commonalities exist, such as the need for the right talent and the right infrastructure. Too much infrastructure can hurt ROI, while too little won’t get the job done. In financial markets, hedge funds typically lead due to their budget and lower regulatory constraints, followed by banks and then institutional asset managers and sovereign funds.
In our experience, we primarily address challenges for hedge funds first, followed by banks and the asset management community. Our engagements vary: sometimes we build and deploy machine learning or AI models, and other times we develop applications to make existing solutions consumable. We call our approach contextual data and technology solutions because it is highly relevant to each institution’s workflow.
TAM: What are some key business imperatives and priorities that CXOs should focus on in the financial services industry? And how do you help them address these priorities?
Deepak Stephen: The first key point is Gen AI, which has been in the market since last year. Its applicability depends on the client segment. In some segments, Gen AI can be applied in its current form because their workflows do not require an audit trail. In these cases, we’ve seen significant adoption. However, in other cases, the impediment is that Gen AI models, especially those like GPT, are not yet explainable. In such situations, CXOs prefer to have verifiable internal code.
Regarding the data deluge, the amount of data has been increasing over the last four or five years. Managing this data and creating a single version of the truth across lines of business is critical.
The third challenge is driving adoption. Change management is particularly difficult in front-office functions. People are accustomed to legacy ways of working, which have been effective and continue to work well, but there’s a need to raise the bar. Compared to a Fortune 500 CXO, the primary problem for financial services CXOs is managing change and adopting data and technology initiatives within their firms.
The pandemic played a role in accelerating cloud adoption, which we have seen across various firms and industries. This has fueled some of the necessary infrastructure, but the financial services sector is still in the early stages of this journey compared to other sectors.
TAM: What are some of the solutions you offer to help people in this industry undergo business transformation?
Deepak Stephen: When looking at their workflow, it’s difficult to generalize across segments, but typically, there is an origination part, an execution part, and a monitoring part. For example, a private equity firm sources talent for origination, executes the deal, and then monitors the portfolio company. Similarly, an investment bank follows these steps.
Let’s consider CRM systems as an example. In a Fortune 100 company, CRM systems have been in place for the last 20 years. However, in some financial industries, the market is fragmented, with over two dozen providers offering CRM systems to private markets, private equity, and private credit. Salesforce is leading, but its implementation requires support and expertise.
Our engagement model usually involves an intense discovery workshop. We map the workflow and explain the changes using a whiteboard, followed by a parallel run and gradual adoption. Maintenance support is essential as people need time to adapt. For example, a Power BI dashboard for a portfolio manager might include multiple data levels, self-servicing, and mobile access. Initially, ensuring features like printable PDFs are available can drive adoption.
Challenges vary among firms. Some are advanced in their quant capabilities, having sorted out data and analytical problems but facing low consumption and adoption. Others have a younger workforce comfortable with data consumption but need to revamp models with new technologies like Gen AI.
Our solution framework, called Acquire, Prepare, Analyze, and Consume, was developed four years ago. It involves application development before and after data consumption. The discovery phase is crucial, dealing with functional leaders and CXOs to get change management sign-off. We integrate solutions step-by-step, understanding unit-level problems and ensuring a smooth transition.
For instance, a bank’s equipment finance unit might use a different CRM linked to their pricing. Changing it overnight isn’t advisable due to the business’s scale. We partner through the journey, providing support until the change management process is complete. Sometimes, we handle the initial stages, and the in-house team takes over later.
TAM: Are there any notable use cases or case studies you would like to share for the benefit of our audience?
Deepak Stephen: ESG adoption is a major focus on Wall Street, especially on the buy side.
Take a US-based asset manager covering a large amount of securities worldwide. They wanted to integrate ESG factors into their fund management. When raising money from investors and deploying it into companies, these companies commit to following the ESG principles defined by the UN. The challenge was determining whether these companies were truly adhering to these principles.
Portfolio managers typically meet with company management quarterly, and if the company is based in regions like Far East Asia, they visit at least twice a year to understand the positives and negatives. For example, let’s consider SDG 5, which focuses on gender diversity. They used AI to assess whether companies were genuinely committed to gender diversity. This involved analyzing two sets of data: internal sources (CSR reports, internal policies, publications, investor relations reports, annual reports) and external sources (news, social media, industry publications).
We created an algorithm that analyzes both data sets and provides a score indicating whether the company is genuinely adhering to its commitments. Experts within the equity domain defined 162 keywords relevant to gender diversity, such as gender-neutral washrooms, returning mothers, and LGBTQ+ policies. This upfront labeling reduces time to market for implementation. Our team validates the algorithm’s performance, ensuring its accuracy.
The impact of this solution is significant. If portfolio companies are not meeting SDG 5 standards, the portfolio management group can either reduce their funds or take corrective action due to their substantial stake. This is just one example of ESG adoption, and many firms across the US and Europe are implementing similar strategies.
Another example involves banks with 15 to 20 years of operations and extensive CRM data. When a capital-raising deal arises, traditional methods involve discussing with investors and taking notes, which works fine for small, regional operations. However, as these banks go global with larger mandates, managing a repository of 10,000 investors becomes challenging.
In such cases, we use simple machine learning algorithms like classification or clustering to identify the most suitable investors for raising debt in a specific market. These algorithms consider factors like sector interest and deal size, leveraging past transactions to predict the likelihood of investors writing a check. This makes pitches more efficient and increases the likelihood of successful deals. We deploy this solution as an application within the bank, enabling them to use it effectively.
TAM: What are some of the key tech trends that are reshaping the global financial services landscape?
Deepak Stephen: To achieve a unified version of the truth, integrating various solutions is essential. For instance, companies typically have separate systems for CRM, HR operations, finance, and front-office functions like trading and risk management. The challenge is managing the data deluge, emerging AI technologies, and the influx of new applications.
While an ideal architecture diagram might show a perfect solution, reaching that ideal state requires navigating through different functions, geographies, and personalities. We are a platform-agnostic provider, meaning we work with the solutions our clients already have. For example, if a client prefers SAS for analytics, we use SAS. If they prefer Python, we use Python. Our approach adapts to the specific challenges our clients are facing.
For mature organizations that have already implemented many solutions, we focus on adoption and change management. However, some organizations are still building their platforms and moving from on-premises to cloud environments for better computing power, security, and cost-effectiveness. In these cases, we partner throughout their value chain. Whether it’s developing new applications, working on the consumption layer, or addressing public-facing solutions, we adapt our approach based on their needs. Our goal is to align with our clients’ skills and objectives, acting as a partner to support and drive their initiatives forward.
TAM: What would some of the key priorities for Acuity Knowledge Partners be in the near future?
Deepak Stephen: From our perspective, attracting and retaining high-quality talent is crucial. Our clients often look for a blend of financial and technical expertise in their staff. Generally, we focus on three key elements: education, toolkit, and domain knowledge. While we typically cover the first two—ensuring our team has the right education and toolkit—understanding front-office capital markets is rarer in India.
We invest heavily in strong internal training programs and spend a lot of time hiring the right people. Additionally, we have established third-party partnerships to upskill our staff. For example, while companies previously focused on just one cloud provider, such as Azure or AWS, we now see more diverse needs due to legacy systems and M&A activities. This means we often need to add new skills related to platforms like GCP.
Acuity Knowledge Partners also partners with external firms for specialized CRM and AI solutions to accelerate our market entry rather than developing these solutions in-house.
Keeping a close watch on macro trends is another priority for Acuity Knowledge Partners. The pace of change varies globally: the US moves quickly, followed by the UK and the rest of Europe, while Asia, including Singapore and Hong Kong, tends to follow these trends later. Staying attuned to these shifts is essential for us to stay relevant and effective. These are the key areas currently occupying our attention.