Deeper Insights

Introduce your business and what you do there.

I’m the co-founder of Marine Judge, which is an AIML cargo claims business. We’re trying to automate the cargo claims business in maritime insurance.

What challenge were you trying to address with Deeper Insights?

Our platform has effective automation capabilities, but we wanted to incorporate AIML to massively scale and speed up that automation.

What was the scope of their involvement?

 Deeper Insights helped us improve that automation so it can make intelligent decisions around cargo claims. That included everything from the ability to import information into the system on whether or not to accept a claim.

Initially, we gave them a brief on the platform and how it works. Then we outlined the three things we wanted to deliver in order to speed up the automation process.

Then we went through how that would look and how the modules would work together. After that, they gave us specific programs to build each of them.

They helped us position us to clients. They were very good at being able to explain the technology, how it worked, and what it was doing. The second bit was building modules so we could actually deliver it.

The third piece was integrating the solution with our platform and making sure that those modules worked well for our platform.

What is the team composition?

It was four people from their team, including Jack (CEO), Eduardo (Operations Manager & Architect), and two of their team members.

How did you come to work with Deeper Insights?

We talked to a number of providers that we found online. We were looking for partners that would help us understand the challenges and various ways to deal with it.

How much have you invested with them?

We’ve spent about £40,000 (approximately $52,000 USD).

What is the status of this engagement?

We started working with them in January 2020 and our partnership is ongoing.

What evidence can you share that demonstrates the impact of the engagement? 

We’ve won three new clients and have been able to roll it out to a number of additional clients. I’d give them five stars for delivering what we needed, and they did even better while working with our clients. For a small business like us, each client is critical, and Deeper Insights really helped in that aspect.

How did Deeper Insights perform from a project management standpoint?

They spent quite a bit of time getting to understand the challenge. They were very good at interacting with it as we were going through that process, so we were going through to make sure that we fully understood where they were and what they were doing. They communicated very well and we enjoyed interacting with them.

What did you find most impressive about them?

Their ability to understand the problem and how it affected the clients stood out. We really needed a partner that understood AIML. Otherwise, it would have been really easy to waste a lot of money.

Are there any areas they could improve?

They’re growing fast, so they’re adding additional expertise. They could be better at explaining how that can be valuable to us. Of course, they delivered everything we wanted but they could broaden their proposition and help us grow.

Do you have any advice for potential customers?

They’re experts in their space and really listen to their clients. New clients can feel comfortable and confident in their partnership with them. They provide good advice about what you need to build and how much it will cost. They really helped us think through what was best for our business, so they were true partners.

Please describe your company and your position there.

I'm the founder of EcoHedge Ltd., we are an Environmental Analytics Start-Up from London, currently with a team of 3.

For what projects/services did your company hire Deeper Insights, and what were your goals?

EcoHedge was looking to assess the viability of replacing a human-based data collection task (specifically sustainability data for UK based retailers) with various Machine-Based web extraction technologies across data repositories, pdfs and web pages. Given the limitations of our existing data collection process, our goals were to:

1. Reduce costs,
2. Reduce time taken to perform data collection,
3. Automatically include new companies/retailers, while maintaining the accuracy of the collected data.

How did you select Deeper Insights and what were the deciding factors?

We selected Deeper Insights through using the internet to research companies within this space and using their website to get in touch.

Within a matter of hours Deeper Insights had responded to our web enquiry and had offered an introductory call to understand our problem and talk through how they might be able to help us achieve our goals.

The. main deciding factor was based on how bought into our goals and project Jack appeared from the outset, it was also helpful that they were willing and able to move extremely quickly to start our work (2 weeks turnaround from initial conversation to project start).

They were also flexible on the payment terms with us which was a nice touch.

Describe the scope of work in detail, including the project steps, key deliverables, and technologies used.

The scope was to carry out a Feasibility Study for the applications of Artificial Intelligence and Data Science techniques in the use of data aggregation, processing, and analysis for EcoHedge. The steps included:

1. Analysing the existing data, processes and systems
2. Identifying issues with existing solution
3. Researching the latest academic or commercial applications in use for this kind of problem
4. Based on research, recommending the best models and method for development (including data requirements, annotation and training) that will give the best results
5. Reviewing additional requirements.
6. Preparing a Project Plan, outlining the approach including investment and time estimates.
7. Presenting the Project Plan to us.

How many people from the vendor's team worked with you, and what were their positions?

We worked with a total of 2 team members from Deeper Insights: Jack Hampson - CEO Marcia Oliveira - Lead Data Scientist

Can you share any measurable outcomes of the project or general feedback about the deliverables?

Deeper Insights were able to meet and exceed our goals, actually offering a solution for a problem that initially had not been within scope of the project;

1. Reduce costs - Deep Insights we able to propose a solution which would reduce our ongoing costs from c. £44 per data collection to c. £1.75 per data collection.
2. Reduce time taken to perform data collection - Deeper Insights we able to propose a solution which reduced time from 30 minutes per data collection to 5 minutes.
3. Automatically include new companies/retailers, while maintaining the accuracy of the collected data. - Deeper Insights successfully proposed a solution for this and also enhanced our name matching process.

Describe their project management style, including communication tools and timelines.

The project was successfully managed through highly responsive communication over email and frequent Zoom-based video calls.

What did you find most impressive or unique about this company?

The unique thing about this company is the access to a founder who is passionate. and really cares about helping you to find the right solution, not push the most expensive one.

For example we were recommended to not use a particular technology as its results would not be reliable for us even though commercially this solution would have been in the interests of Deeper Insights.

Are there any areas for improvement or something they could have done differently?

Overall I am delighted with our recent engagement with Deeper Insights. As such I do not have any particular areas for improvement. One small thing which might be worth considering is providing the initial scoping questionnaire in a more user friendly format.

I struggled a little with the excel document and imagine there are better web based form alternatives, whereby this process could be streamlined for both parties. This would also cut down the (albeit small) risk of a data leak through mis-sending the excel.

Please describe your company and your position there.

I am Head of Data at DACS. An organisation established by artists for artists, DACS is a not-for-profit visual artists’ rights management organisation. The organisation is passionate about transforming the financial landscape for visual artists through innovative new products and services, and act as a trusted broker for 100,000 artists worldwide. Founded over 30 years ago, DACS is a flagship organisation that campaigns for artists’ rights, championing their sustained and vital contribution to the creative economy. We collect and distribute royalties to visual artists and their estates through Payback, Artist's Resale Right, Copyright Licensing and Artimage. Since we were founded in 1984, we have paid over £100 million in royalties to artists and their estates – a significant source of income supporting artists’ livelihoods, their practice and legacy. In 2018, we paid £18 million in royalties to artists and estates.

For what projects/services did your company hire Skim Technologies?

Keeping a track of the volume of transactions made on artwork produced by artist we represent (members) and those that are not yet representing is a process that has been delivered manually for years. This process is used to forecast qualifying sales based on a range of metrics and also audit the information provided by auction houses and other Art Market Professionals. Skim Technologies was engaged to create a Machine Learning solution that automates this process end-to-end.

What were your goals for this project?

Valuable time invested in sourcing forecast information for ARR qualifying sales from various websites will no longer be required. ​ Generation of potential leads to boost recruitment drive and ultimately increase revenue growth.​

How did you select this vendor?

Skim Technologies was engaged on the project following on from referrals

Describe the project and the services they provided in detail.

The solution/tool comprise of Machine Learning models, Databases, Modelling APIs, Visualisation with Power BI and a third-party product (Prodigy) used for active/supervised learning​ Fully automated workflow with AI comprising of an ensemble of several models performing a range of tasks​ The models are currently being trained through supervised machine learning with outputs generated every 24 hours​ With the completion of phase 2, the question on the hosting location is yet to be confirmed however, the recommendation is that DACS subscribes to the hosting service provided by the developer as this will be more cost effective plus there are no in-house capabilities to maintain the solution.

What was the team composition?

1 x Project Lead 2 x Data Scientists 1 x DevOps

Can you share any information that demonstrates the impact that this project has had on your business?

DACS is now leveraging on Artificial Intelligence capabilities to improve internal processes for ARR forecasting and increase revenue generation capacity.​

How was project management arranged and how effective was it?

Skim Technologies led on the management of the project with frequent and ad-hoc engagements with stakeholders at DACS

What did you find most impressive about this company?

Apart from the fact the Skim Technologies team are absolutely brilliant and knowledgeable in their space, they quickly developed the requisite knowledge about our industry and internal processes.

Are there any areas for improvement?

No. 

Introduce your business and what you do there.

I am a product manager at a data analytics company. We provide alternative data and metrics to analyze financial security. We use a unique process of analysis to determine what the main macro drivers are of any particular asset class.

What challenge were you trying to address with Skim Technologies?

We hired Skim Technologies to develop an exploratory prototype of a chatbot that would allow our clients to receive actionable investment analysis quickly and efficiently.

What was the scope of their involvement?

We relied on Skim Technologies to build a website chatbot that could be used to respond to three central interactions. We wanted the chatbot to serve users who wanted to learn more about a trade or a stock; users who, for example, thought that the price of oil was going up and wanted to know about the best expression of that trade; and users who wanted to know about the best opportunities on the market, which explores what stocks are currently over- or undervalued on our company’s radar.

To facilitate those interactions, we had to build a database of questions that would map to one of those requests. We had to come up with all of the different ways that users might ask what is essentially the same question. Further, users might refer to crude oil, for example, by any number of names. We had to build out that database as well to make it as user-friendly as possible.

What is the team composition?

I worked primarily with Jack (CEO, Skim Technologies) and a pair of coders and data scientists.

How did you come to work with Skim Technologies?

We met them through one of our partners that is involved in developing our platform API.

How much have you invested with them?

We’ve spent between $10,000–$40,000.

What is the status of this engagement?

We worked together from January–March 2018. The chatbot was an exploratory project that we have had to deprioritize in favor of our core products. Our ultimate goal is to build it out into an actual product eventually.

What evidence can you share that demonstrates the impact of the engagement?

Skim Technologies delivered a functional but limited chatbot that we use internally to show users, investors, and clients the power of our company in a unique way. The chatbot is not a product that we sell, but it has served us well as a marketing tool to this point. It’s sparked interest and has been a conversation starter for us when we’re pitching to clients. That chatbot is undoubtedly useful, but the real value for us will come down the line. It has huge potential.

How did Skim Technologies perform from a project management standpoint?

They fulfilled the brief and delivered exactly what we wanted. They were easy to work with—both through emails and face-to-face meetings.

What did you find most impressive about them?

We hired them for their expertise, and they delivered it.

Are there any areas they could improve?

No, nothing comes to mind. Overall, it was a productive experience. I believe that if we pick this project up again in the next few months, we’ll be speaking with them. They did a good job originally of understanding our business and our requirements.

Do you have any advice for potential customers?

As with any engagement, it’s critical to be clear about what the end goal is. We also found it valuable to have constant communication with the developers.

Introduce your business and what you do there.

I’m the content and marketing officer for an app which has been developed by a charity to support women after their hospital-based breast cancer treatment.

What challenge were you trying to address with Skim Technologies?

As a charity, we know that finishing hospital-based treatment for breast cancer can be a very difficult time for many women. There’s an expectation that everything will go back to normal, but women often instead experience a decline in their emotional well-being. We created an application that provides tips, information, and inspiration on easy-to-use flashcards to help women adjust to their "new normal." Our content covers a variety of topics, ranging from symptoms and concerns to mindfulness and exercise.

Our internal content team was dedicating a lot of time to scanning forums, blog posts, and various healthcare websites for information to populate the app. To increase our efficiency, we decided to look for an outside vendor to help build a CMS for our platform.

What was the scope of their involvement?

Skim Technologies helped us develop a CMS to find the right information for our app more efficiently. We can access it on any desktop, and it’s connected to the backend of the platform. It allows our content team to create diverse content and gather information from around the web.

The Skim Technologies team implemented an API that connects their machine learning model with the new CMS. The model consistently searches the internet for sources of information that are potentially relevant to our organization and pulls it into the CMS. Our internal team can view a summary of the content and either accept, reject, or hide each item. This scoring system helps Skim Technologies' machine learning model better understand the type of information we're looking for. As the model collects more data on the way we interact with the content, it will start aggregating more relevant information. Eventually, it will become a more automated process, minimizing the need for editing before publishing information on the platform. We’re still working with the Skim Technologies team to refine and enhance the process.

We're also working to personalize our content for each user, and Skim Technologies will help us collect data on different ways users interact with the information on the app. Based on that feedback, we’ll be able to provide each individual with content tailored specifically to their needs.

What is the team composition?

We have one direct point of contact who manages a team of data scientists.

How did you come to work with Skim Technologies?

Our digital innovation manager, who developed the app, was familiar with Skim Technologies and their service offering. She reached out to them when we were applying for a funding grant to make sure our project was sustainable over time.

What is the status of this engagement?

We started working with them in May 2017, and the project is ongoing.

What evidence can you share that demonstrates the impact of the engagement?

Skim Technologies continues to improve and refine the system. Over time, the content has become more and more relevant. They understand and properly address all of our feedback, which has been great. So far, we haven’t experienced any bugs or crashes, and if there’s a small glitch, it’s handled promptly. Overall, it’s been a smooth experience. We've been impressed with the quality of their work and their services.

​​How did Skim Technologies perform from a project management standpoint?

They’ve been excellent. We speak every two weeks, and everyone involved in the project attends. We discuss our status, and any progress updates or upcoming challenges. It’s been an interactive process, incorporating feedback in each step. In addition to those meetings, we track the project using Trello.

What did you find most impressive about them?

They’re professional in receiving feedback and responding to it quickly. Open and frequent communication has been key to our relationship.

Are there any areas they could improve?

No, I can’t think of anything.

Any advice for potential customers?

There needs to be an open dialogue regarding project goals. We identified our primary goal at the beginning—to find information quickly and efficiently for our content app, and alleviate some of the capacity demands on our internal team. By involving Skim Technologies in discussing long-term aims, they were able to help us identify solutions and structure the full scope of the project.