Target Screening for AI Inferencing Investment – MENA Strategy Fund
Mock Case Interview
Writing this blog took considerable effort. If you enjoyed it, please consider subscribing. Feel free to leave a comment if you'd like me to cover a specific industry or function in the next post.
Additionally, I run a digital e-commerce platform for business education on Gumroad: https://karandhir8.gumroad.com. I also mentor students on https://topmate.io/karan_dhir10
Interviewer:
Our client is a UAE-based investment fund backed by sovereign capital. They’ve made infrastructure plays in cloud computing, smart city platforms, and data centers. Now, they want to build leadership in AI inferencing infrastructure in the MENA region. They’ve asked us to identify 1–2 global companies in this space for acquisition or investment.
How would you structure your approach to this target screening?
Candidate:
Thanks for the context. I’d break the problem down into four broad areas:
Clarify the client’s strategic intent – What outcomes are they targeting with this investment? Is it access to IP, customers, infrastructure, or talent?
Define screening dimensions – I’d build a framework to evaluate companies based on market, technology, financial, and strategic fit.
Shortlist promising candidates – Using filters from public/private data sources and applying the framework.
Recommend 1–2 targets – Based on scoring and fit, then outline next steps (e.g., due diligence, integration).
Does that sound like a good starting point?
Interviewer:
That makes sense. Let’s focus first on what dimensions you would use to screen companies. What are the key criteria?
Candidate:
Sure, I’d group the screening dimensions into five buckets:
Market Attractiveness
Size and growth of AI inferencing (especially in edge/cloud deployments)
Demand in verticals aligned with MENA (e.g., smart cities, logistics, public sector)
Local infrastructure gaps (e.g., GPU shortages, latency bottlenecks)
Technology Differentiation
IP around model compression, quantization, or low-latency serving
Compatibility with edge/cloud/hybrid deployments
Developer APIs and model portability (e.g., ONNX, TensorRT, etc.)
Partnerships with hyperscalers, cloud providers, or telcos
Financial Fundamentals
Stage: Series A–C (with sub-$200M valuation)
Revenue run rate, growth trajectory, gross margin
Burn rate and funding runway
Strategic Fit for MENA
Potential to localize stack in UAE data centers
Compatibility with existing fund assets (cloud, 5G, etc.)
Talent mobility (e.g., proximity to Europe/Israel for ease of expansion)
Readiness for GCC regulation and sovereign data localization laws
Risk Factors
Export controls on hardware/software (esp. US-origin chips)
IP litigation or exclusivity risks
Founder dependence or early customer concentration
Interviewer:
That’s a thorough framework. Let’s say you’ve applied this and come up with 3 companies. Walk me through how you would select the final target.
Target Evaluation (Sample Candidates)
Candidate:
Sure, I’ll present three hypothetical finalists and compare them using a scoring matrix based on the five buckets:
I would compare them across six key lenses:
1. Technology Differentiation:
Deci.ai wins on pure software play for edge and model compression. Ideal for smart city/public safety use cases.
NeuReality offers custom silicon—future-proof, but heavy to deploy.
2. Scalability in UAE:
Inferless enables a fast roll-out in cloud zones via APIs — great for lightweight deployments.
Deci.ai can scale across logistics, border control, and mobile inference.
3. Integration Time & Complexity:
Inferless is SaaS-based — quickest to integrate.
NeuReality would require full hardware-software stack integration (slower GTM).
4. Valuation & Deal Dynamics:
Deci.ai is growing fast — acquisition possible now before it becomes too expensive.
Inferless may accept a joint venture or Series B lead investment.
NeuReality may demand higher upfront CapEx.
5. Regulatory & IP Risk:
Deci.ai and NeuReality are Israel/EU-based — less export risk than US-based targets.
Inferless may face hurdles with US-origin GPU exports or changing policies.
6. Strategic Fit with Sovereign Vision:
Deci.ai aligns with smart infrastructure and edge goals.
NeuReality matches UAE's desire for deeptech leadership and chip localization.
Final Recommendation
Candidate
Primary Target: Deci.ai – via acquisition or majority JV
Brings best mix of IP, deployability, and edge readiness
Aligns with UAE’s smart city and edge AI goals
Israel-UAE relations enable smoother M&A and tech transfer
Secondary Option: Strategic Investment in Inferless
Allows quick presence in inferencing SaaS market
Cloud-neutral — a hedge against hyperscaler lock-in
Option to acquire later based on performance
Interviewer: Good. What would be your follow-up actions post-selection?
Follow-Up Action Plan
Candidate
Initiate 60-day due diligence
IP/legal validation
Financial and technical deep-dive
Founders' vision and alignment
Integration Blueprint
Host stack in UAE zones (e.g., KIZAD, Hub71)
Partner with government on regulatory certification
Build AI R&D center with target company
University Partnerships
Joint labs with Khalifa University, MBZUAI
Local model tuning and student internships
Narrative Building
Frame UAE as the AI inferencing hub for emerging markets
Position fund as an alternative to hyperscaler-led AI infra
Bonus Insight (If Asked): How is inferencing different from training in AI?
Candidate
More latency-sensitive (e.g., in cameras, sensors)
Cheaper and more frequent
Expected to form >90% of AI workloads by 2028
Hence, inferencing infrastructure is the last-mile bottleneck — and the biggest commercial opportunity.