Windfall Geotek (WIN) - Technical Analysis

Windfall Geotek (formerly known as Albert Mining) is a Canadian corporation offering a proven and industry-leading digital platform leveraging Artificial Intelligence (AI) technologies to significantly improve outcomes in the exploration, development, operations, and financing of mining projects. The principal market encompasses the global resource mining industry that comprises virtually all forms of mineralization including oil and gas exploration. 

Merlin-Marr Johnson caught up with Simran Kamboj, President, and CTO, along with Nathan Tribble, Director, Windfall Geotek. 

Mr. Kamboj brings over a decade of technical engineering experience in machine learning and artificial intelligence. Recently, he led a team at Lixar, a division of BDO, where he worked on AI solutions for mining, and electrification projects. Prior to his tenure at Lixar, Mr. Kamboj led product management at Patriot One Technologies (TSX: PAT, OTC: PTOTF). His educational credentials include a degree in biomedical engineering from Ryerson University. 

Mr. Tribble has over 14 years of professional experience in exploration and mining, with a particular focus on gold and base metal exploration and project evaluation. He currently holds the position of the Vice President of Exploration at Gatling Exploration Inc. His past experience includes Senior Principal Geologist for Sprott Mining, Senior Geologist for Bonterra Resources, Jerritt Canyon Gold, Kerr Mines, Northern Gold, Lake Shore Gold, and Vale Inco. He sits on multiple boards affiliated with the mining industry. His educational credentials include a Bachelor of Science degree in Geology from Laurentian University. He is registered as a Professional Geoscientist in Ontario.

Company Overview

Windfall Geotek is a corporation that provides AI-powered tools to assist in the discovery, exploration, development, and financing of mining operations. The company was founded in 1996 and is headquartered in Canada. SIMACT Alliance Copper Gold Inc and Tropic Diamonds Inc. are the company's subsidiaries. The company is listed on the Toronto Stock Exchange (TSX-V: WIN), the OTC Markets (OTCQD: WINKF), and the Frankfurt Stock Exchange (FWB: L7C2). Since 2004, the company has added value to over 30 client discoveries and more than 80 target generation projects around the world. 

Windfall Geotek is a market leader in machine learning and AI (Artificial Intelligence). The company uses machine learning and AI in early exploration. The company utilises all the data that a geologist would use but manages to cut through significantly more noise to demonstrate anomalies in mineralization. The company has been around for 15 years and has proven itself on multiple occasions. 

AI in Mining

Windfall Geotek has been grounded in a lot of research and development that has taken a long time to prove. While other industries have implemented AI for years, the mining industry has taken longer to adopt the emerging technology. One major difference when implementing AI technologies in mining is that the company has to wait for the whole cycle. It is expected to provide a target, following which the operator has to be able to conduct exploration and go through all the stages to get to an operating mine. Notably, Windfall Geotek is the first AI company that has an operating mine which is a direct result of its work. The company intends to become an invaluable tool that every geologist uses in early exploration programs. 

It is interesting to note that Windfall Geotek has been around even before the iPhone was launched. The company’s official spin-off took place 15 years ago. This gives an idea of the scale at which mining industries need to bring useful technologies to the domain. Contrary to popular belief, this won’t cause the loss of jobs, but instead, provide a way to modernise the process. 

Windfall Geotek’s technology has stood the test of time. Generally, in mining operations, there are cycles of cash flows and cash crunches. The company is majorly focused on generating an outcome for a business through the implementation of AI. The business outcomes help Windfall Geotek in attracting other mining companies that would be interested in implementing these technologies. 

Company Workflow

Most AI technologies are ideally suited and highly rated for a data-rich environment. Despite being a dependency on AI, this is stage one of the operation. In order to progress to the next step, more data is needed, this can become a limiting factor for AI. 

However, in the case of brownfield and near-mine operations, it is often seen that companies can take advantage of the technology with a small amount of available data. AI can also work in greenfield operations where the asset is based next to a mine. However, mining companies with brownfield operations have an advantage. Here, the existing data can be used to gain new insights. 

The company does not believe in making empty promises, as some of the large players in the market have been burned by AI in the past. This was due to the early market conditions and the hype relative to the outcomes that could be delivered.  It is important to note that the data comes with certain limitations. Windfall Geotek provides a guardrail for all its clients before asking them to turn over the data for review. 

The main challenge for the company is to not burn its reputation. This is because AI isn’t a standalone solution, but a part of a larger stack of technologies that early exploration geologists would use. The company conducts a free valuation by reviewing the client’s existing data, but it won’t proceed ahead unless value can be created. The company is looking to develop ways in which it can work in lower-data environments with a higher degree of certainty. The company has already started implementing this up to a certain extent. 

For instance: Playfair Mining Limited was a client with limited available data and no drill data. Windfall Geotek was still able to help the company. Playfair was in a very early phase of exploration and Windfall managed to provide it with 22 targets. Upon soil sampling and geology inspection, 16 of these 22 targets produced net positive results. Although it is important to note that low-data environments are prone to a higher number of false positives. Windfall believes that companies with brownfield projects have a massive opportunity to utilise the technology. 

Traditionally in an exploration project, a company would look at coincident anomalies, which are then used to build a GIS data set. It builds layers of information including geological information, geographical data, and a whole range of different parameters. In addition, the company also looks at remote sensing images, photographs, and LIDAR (Light Detection and Ranging), along with the geophysics that goes into the rock geochemistry, the soil geochemistry, the faults, and fractures. Normally, an experienced geologist or a geoscientist will look through the data set and pull out targets based on experience. Windfall’s AI is designed to look through millions of data points with all the different layers combined. This gives the AI an advantage as humans have limited capacity to find correlations between multiple layers and extensive datasets. 

Data and geologists have an interesting relationship. A property’s dataset includes geophysics, magnetic signatures, geochemical signatures, faults, spatial data, and more. Using AI is similar to having a table of geologists and geophysicists to obtain all the maps and interpretations. However, AI has the capability to go one step further by compressing all the data at once, looking for positive patterns and signatures, cloning them, and identifying unexplored areas with similar signatures. There needs to be a specific amount of data that is required for processing, which can result in beautiful signatures obtained from the cloning of nearby positive correlations of a magnetic signature that is associated with a fault. This signature can then be relayed to a different unexplored area. 

Once the signatures are delivered across different areas, a round table discussion is conducted between the geologists, geophysicists, and the exploration manager to interpret the resulting data. This helps the development and setup for the next round of exploration. One example of this workflow is Metanor Resources’ Bachelor mine area in Norway. The company had only conducted soil sampling in the early days. Following this, the company drilled a target and defined a new resource in its underground bulk sampling. This project went from being completed untouched to production as a result of Windfall’s interpretation and target generation. 

Windfall Geotek is proud of its past achievements. The company is looking to work with different groups and is focused on helping teams get discoveries. It seeks to accelerate programs in a sustainable manner. The company’s technology is intrinsic in nature, and it serves as a big step in the evolution of the mining industry.

The Moroy Zone Target

Windfall Geotek provides the output to its clients in the form of high similarity and low similarity targets that have a chance of mineralization. The company shared the resultant data from a 2010 project that was based in the Metanor’s Bachelor Lake mine area. Here, the red and pink dots represent the higher similarity target while the green dots are data points that represent lower similarity targets. The driving force behind the similarity model is heat map signatures. The ones in yellow are above a certain grade, while the ones in green are indicative of a certain grade. 

Interestingly, Nathan Tribble, the company’s Director was a client before he joined the company. Following the conclusion of the work at Metanor’s Bachelor asset in 2010, the project was filed away. However, in 2017, Nathan was working with the exploration team on a different project called the Bari, which is south of the Metanor asset. 

He was going through the database in an attempt to generate some new targets and during this process, Nathan stumbled across some completed AI work. Following this, Windfall’s team demonstrated that the 2015 interpretations were a result of drilling based on AI work without the use of heat signatures. The AI helped validate the Moroy target which is also called Target A. In 2015, 10g of 26m was drilled even before the data sets were utilised. 

Windfall Geotek had validated the project without even knowing it. The company had also worked on the western side of the project. Here, it had 3 samples on the surface in 2010, which was followed up by limited exploration. During this time, the company went through managerial changes that led to the loss of data. During this time, AI was put out of commission. When Nathan joined the company in 2017, he brought the AI back into use. Nathan pushed to explore the 26m of 10g drill hole further. Additional drilling was conducted at the Moroy target, which was followed by resource delineation.

The Bachelor mine is located south of Target A (the Moroy target). This area has a strong presence of high similarity targets, that are represented by pink dots. The client drifted underground from the lower level. Currently, mining and bulk sampling is being done in the Moroy zone. In essence, Windfall Geotek generated the target 5 years before the asset was actually drilled and mined. 

Generally, it takes a company about 10 years to go from exploration to mining. It took Metanor Mining Inc. 5 years to drill the area. In March 2021, the company decided to drill the other side and targets, hitting multiple areas of widespread low-grade mineralization. 

Superimposing a geology map on top leads to the realisation that all the zones are bounding on an intrusive granite unit. The mineralization here tends to hug the edge of the intrusion. This finding was picked by Windfall’s AI, which led to multiple validations all over the property. This demonstrates that the work done by the company 12 years ago still holds true, even up to the 2021 drill program. 

Metanor is currently conducting surface exploration at Zone B. At Zone A, the company has drilled the top portion down to a certain level. At the same time, Metanor was planning to drift over and use the underground infrastructure for drilling from below. Drilling in this fashion is comparatively inexpensive. The company has started delineating the resource. It anticipates that the Bachelor target (Zone B) and the Moroy target (Zone A) meet at the bottom. As a result, the company has drifted over and initiated bulk sampling at Moroy to test consistency in grade. Following this, the supply would be sent to the mill. 

The mill’s tailing facility is located on the right of Zone A. The company never followed up on the area which was located on the proximal edge proximal. This region hosts significant mineralization and is also the location of Windfall’s target. Based on the depth of the survey, Windfall Geotek can project a geophysical survey or anomaly. In most cases, the company can go up to 200m below the surface. However, this is largely dependent on the survey as the depth is limited by the available data. 

Target Generation

Windfall Geotek currently generates targets in two dimensions. This is due to the limitations of available data beyond a 200m depth. This is a limitation of geoscience and magnetic surveys. The AI is dependent on the data input, and as a result, is bound by the same limitations. The company is very careful with data inputs as any incorrect data can lead to an odd anomaly. 

At present, 200m of the magnetic survey seems to be the limiting factor. However, if a deeper penetrating magnetic survey is utilised, the company can look at the relationship between the drill data and the magnetic survey, which can be a game-changer. The AI can use all the available data to pull out advanced correlations. 

Mr. Simran Kamboj, the company’s President, and Chief Technology Officer was brought into Windfall Geotek due to his extensive experience in machine learning and AI. He was introduced to AI through Dr. Geoffrey Hinton, a well-renowned authority on neural networks. Dr. Geoffrey sparked Simran’s interest in the sector 7-8 years ago. As a result, Simran switched his career from finance to the energy sector. 

The company has employed extensive R&D (Research and Development) to further improve its AI and stack of technologies. This has enabled the company to cut down processing times from 12-16 weeks to around 2 weeks. It is important to note that the processing times are also dependent on the data sets and the data packaging by the client. Reducing the processing time has allowed Windfall to significantly increase its capacity by building systems at scale. Notably, a company team member has built one of the largest AWS (Amazon Web Services) databases hosted in Canada. He is in the same team as the people who have built real-time systems for energy trading. 

The team has collectively built a system that gives the company leverage not only on the tech side but on the business front as well. For instance, a client approached Windfall Geotek right before winter. The client had a last-minute interest to explore targets just before the first snowfall. Windfall was able to provide results within a week, enabling the client to explore the areas, conduct rock sampling, and find anomalies. This led to a significant acceleration of the program, which would otherwise require waiting until springtime. Shorter processing times provide companies a massive advantage over the competition. Furthermore, AI offers an unbiased interpretation of that when compared to human-based analysis. 

Windfall Geotek has also worked on Canada Nickel’s project. Notably, the project featured 2 land packages. Windfall took out the middle area as the power line here skewed the magnetic data. Canada Nickel was looking to get the existing data validated, while also looking for any missed details. Windfall Geotek crunched the existing data, including all the layers and the regional data. Since the company isn’t bound by Canada Nickel’s proprietary data, it can scan a larger area. The land packages were located close to Timmins, where Windfall has extensive open source and public data. The data processing led to the identification of high priority targets, which upon drilling led to the discovery of one of Canada’s best high-tonnage, low-grade deposits. Interestingly, It took Canada Nickel between 1-1.5 years from acquiring the targets and putting out a resource. The company’s geology team had internal models which had significant overlaps with Windfall’s models in multiple areas. 

Business Model

Windfall Geotek’s business model is three-fold. The first part of the model is to generate value for customers in the immediate term, generating cash flow. At the same time, the company is building an asset portfolio including royalties that have actual stakes in properties. It is also open to optioning off the properties that have been found using public data to a client who is interested in exploration. This helps increase the chances of discoveries. 

From an investor’s perspective, Windfall Geotek offers a long-term holding opportunity on metals and the junior market. The company anticipates that its asset portfolio will pay out in the mid-term and the long term. It is looking to raise sufficient capital so that it can become cash flow neutral. The company’s current burn rate is about CAD$1.5M-$1.6M on a yearly basis. The burn rate for 2021 was higher because the company had made investments in scaling up operations. The burn rate for 2022 is expected to be similar to or lower than the previous year. The company is looking to generate sufficient revenue to ensure that it doesn’t need to raise additional capital. 

Target Interpretation and Delivery

Windfall Geotek has multiple geologists on the team. When working on a project, camp, or area, the geologists come up with individual interpretations and then conduct meetings to discuss the findings. This enables the company to have confidence in its interpretation and AI work and avoids the generation of a signature in the end. This is also done to ensure that the targets given to the clients line up with the structures and are logical from the perspective of a magnetic signature. This is followed by the interpretation process from both the client and the company. 

The final deliverable is in an ArcGIS compatible format which makes it easy to insert into any exploration program. The areas represented by red and pink are the high similarity areas. The client can see all the targets that Windfall considers are similar to 99% of the style of mineralization that was being originally sought. The mineralization can also be dragged down to 65%-70%, which is represented by yellow and green pixels. The company’s approach is to look at the broader scale and dial back before drilling is commenced. Interpretations are shared between the geologists and an agreement is reached before a client meeting is set up. 

During a client presentation, the company explains the generated results. These results are mathematical models that aren’t binary in nature. The outcome is based on modelling. The company puts forward its own interpretations and targets before viewing the client’s targets. This helps take out bias from the equation and actually helps instil confidence and validation when the targets line up. Following this, the reasoning behind high-priority targets is explained based on a structural point of view, and geological and geophysical positioning that wasn't taken into account by the client. 

Target 2022 and Beyond

Ideally, Windfall Geotek is looking to evenly split resources between its own target generation and portfolio development. However, earlier this year, the company had more client work than internal work as a lot of clients were looking to get an outcome before the spring season. The clients wanted some time to review the outcome, and as a result, the company had to process it upfront. Windfall Geotek is also focused on its own projects. It anticipates that sufficient revenue can be generated this year to conduct both client work and internal work. 

Notably, the internal work can be highly rewarding at times. For instance, if the company runs interpretation on large mining camps and comes up with good targets, it is later found that all the land packages are already taken. In other cases, the company utilises its existing claims which can be optioned to a third party. Here, the company offers a third party to buy the target and provides a scan for its project as well. This ends up being mutually beneficial for both parties.

When it comes to internal projects, the company may redo the work on existing projects multiple times and compare the outcomes. This analysis enables the company to stake some ground or option it to a third party. This process can take up to a year to come to fruition. This is because when the claims are staked, there are multiple parties viewing each of the claims for a potential option. Other factors such as capital availability, the market cycle, and gold pricing come into play which can result in a longer process. Internal work offers Windfall more control over the data. The outcomes of the labour may not be always seen as the company prioritises confidence in staking claims before disclosing them to the market. 

The company has seen an influx of return clients who have been highly satisfied with past work. These clients are often looking to re-hire Windfall Geotek for their next project. Windfall Geotek seeks to become the go-to AI company for the entire industry at some point in the future. Interestingly, some of the repeat clients worked with the company years ago and decided to appoint Windfall for the next project. 

To find out more, go to the Windfall Geotek website