Disrupting Commodity Markets Through Vertical Integration and Engineering

In this episode of What is Your Problem?, Shinkei CEO Saif Khawaja explains a major bottleneck in industrial food systems: the gap between ethical goals and actual operations. By bringing the Japanese ikijime technique, traditionally a manual process, into industrial fishing, Khawaja is doing more than improving animal welfare. He is forcing a change across the entire supply chain. His work shows that when an industry uses commodity status to mask declining quality, the best way to disrupt it is through vertical integration rather than small improvements. Khawaja uses engineering to solve ethical problems, offering a guide for anyone building a premium brand in a market that favors volume over value.

The hidden cost of commodity efficiency

Most industrial fishing operations focus on throughput. As Khawaja points out, the current standard involves pulling fish from the water and letting them suffocate. This is an ethical failure and a significant operational inefficiency. By failing to process fish immediately, the industry accepts a rapid onset of rigor mortis, usually within two to four hours, which causes biological breakdown and shortens shelf life.

Khawaja’s system uses computer vision to identify fish species and automate the ikijime process, extending this timeline from four hours to about 35 hours. This is a systems level intervention. By solving the humane problem at the point of catch, he also solves a logistics problem, creating a product that stays fresh much longer.

I wish I had something more romantic than thinking why not solve the problem directly again, which is an engineering response and fix the root cause, which is build something that can actually do that process.

-- Saif Khawaja

Why the obvious fix makes things worse

A revealing part of the conversation is the discussion about stunning fish with electricity. Conventional wisdom suggests that adding a stunning step is the logical way to ensure immediate loss of sensibility. However, Khawaja explains why this creates a downstream disaster. Because wild fish vary in size and species, a single voltage setting is impossible to calibrate. If the voltage is too high, it destroys the meat. If it is too low, it causes further suffering.

This highlights a trap in systems thinking: the obvious solution often introduces new variables that the system cannot handle. Khawaja’s team ignored the conventional stunning step and focused on speed. By reducing the cycle time to a few seconds, they achieve the desired outcome without the collateral damage of electrical variability.

The 18 month payoff of vertical integration

Khawaja initially tried to sell his machines as a standalone product, a robotics as a service model. It failed. The market did not understand the value because the industry had no language for process quality in fish. To fix this, Khawaja did what most founders avoid: he vertically integrated.

He built a brand, Ceremony, and bought a processing plant. This was a high friction, capital intensive move that most would advise against. Yet, it was the only way to prove the value to skeptical retailers.

We originally actually tried to sell the technology. We didn't really have that as, like we didn't get the response we wanted... unless we did ourselves, people would just not understand the value for what we were recreating.

-- Saif Khawaja

By controlling the entire chain, he forces the market to accept a new standard. He is not just selling a machine; he is selling a different category of fish. This is a competitive moat: doing the hard, unglamorous work of building the infrastructure that others are too impatient to construct.

Key action items

• Audit your commodity processes: Identify where your industry accepts low quality or high waste simply because that is how it has always been done. (Immediate)
• Test for hidden utility: If you are making a product better, quantify the secondary benefits, such as shelf life or durability, that your customers might care about more than your initial value proposition. (Over the next quarter)
• Avoid obvious feature bloat: When solving a problem, resist the urge to add safety or compliance features that introduce new, uncontrollable variables. Focus on speed and simplicity first. (Ongoing)
• Control the narrative through integration: If the market does not understand your innovation, stop trying to sell the tool and start selling the end result. Build the vertical stack necessary to prove your claims. (12 to 18 months)
• Design for variation: If your system fails at scale, it is likely because you are designing for a perfect input. Use computer vision or modular logic to handle the messy, real world variance of your inputs. (6 to 12 months)
• Build the third party yourself: If no certification exists for your new standard, create the data and the audits yourself. Become the benchmark that others eventually have to follow. (18+ months)