You want to make a plant-based dish. We take it from formulation, ingredient selection, cooking condition and packaging through to production support. Items we have developed include sweet and sour plant-based chicken, vegan chilli sauce, mushroom soup, curry rice, vegan bolognese and vegan dumplings, as well as plant-based bacon, cheese and egg.
Plant-based dishes we have developed
These are dishes we have actually developed. Each went through formulation, ingredient selection, cooking condition, packaging and production support. You arrive with an idea and leave with a product that can be manufactured and listed.






Other categories we handle
- Plant-based alternativesPlant-based bacon, plant cheese, plant egg
- Chinese plant-based dishesMapo tofu, salt-and-pepper bites, sweet and sour dishes
- Sauces and basesVegan chilli sauce, vegan bolognese, seasoning bases
- Soups and brothsVegan mushroom soups and plant-based broths
- Dough items and dim sumVegan dumplings with variable fillings, dough items
- Ready meals and retortCurry rice and ambient, chilled or frozen retort meals
What you bring, what we handle
You do not need to understand food processing. Describe what you want to make and the technical judgements are ours. But a few decisions are only yours to make, and they set the direction of the whole project.
- You bring: the item and the flavour directionA concept, a family recipe, or a reference product on the market
- You decide: vegetarian class and target channelFully vegan, ovo-lacto or allium-free, and where it sells
- We handle: formulation and ingredient selectionFlavour, texture, cost and supply stability together
- We handle: cooking condition and scale-upTurning a taste that works small into a specification a line can run
- We handle: packaging and labellingPack selection, label review and nutrition panel calculation
- We handle: line configuration and productionConfiguring a suitable partner line and owning the quality standard and schedule
Flavour depth: two layers vanish at once
Animal ingredients supply two things in a dish, and usually neither is noticed. The first is savouriness from free amino acids and nucleotides. The second is the aroma compounds fat generates during cooking.
Remove the meat and both layers go at once. What the eater registers is not the absence of a meaty note but that the taste is thin, because what was lost is depth on the palate and layers in the nose, not one specific flavour.
Rebuilding it involves fermented ingredients such as soy sauce, miso and yeast extract, the natural umami compounds in mushrooms and seaweed, controlled Maillard development, and the choice and ratio of plant fats. None of it is solved by adding one seasoning; it is formulation work.
The difficulty is that each of those carries its own signature. Too much yeast extract reads fermented, too much mushroom extract reads earthy, and pushed far enough it simply becomes a different dish. Plant-based sampling therefore needs more rounds than an equivalent non-vegetarian product.
Texture: a problem that appears after the kill step
Plant proteins behave differently from animal proteins under heat and pressure. After retort, structure that had bite can turn loose or mealy, and that is invisible in a sample made before the kill step.
This is why a sample that tastes right in a lab or a kitchen does not mean it survives the line heat curve. Industrial ramp rate, hold time and cooling all differ from small-scale trials, and texture is sensitive to all three.
So the formula is designed together with the process condition, and validated with a pilot run on the actual line. This step cannot be extrapolated from small samples. Getting it wrong costs a full production run with the wrong texture.
The frozen route has its own set: weeping and texture change after thawing depend on freezing rate and temperature fluctuation in storage, not only on the formula. That detail is on the frozen food page.
Vegetarian class: one word changes the line
Fully vegan, ovo, lacto, ovo-lacto and allium-free are distinct label classes, and the class decides more than the formula. It decides the segregation and sanitation validation the line has to demonstrate.
A line that also runs animal products may not satisfy the shared-line requirements for a fully vegan claim. Not because the equipment cannot be cleaned, but because proving that every changeover was cleaned needs validated procedures and records, and not every plant has built that.
A product containing allium can be excluded from vegetarian foodservice channels on class grounds. This matters particularly in the Taiwan market, where the allium-free and fully vegan audiences do not entirely overlap.
So the vegetarian class is a mandatory question at scoping, not a checkbox at packaging design. The answer narrows the usable lines, which in turn affects capacity, batch size and cost.
The protein source decides everything after it
Plant protein is not one material. Soy, pea, wheat gluten, rice and mycoprotein differ substantially in flavour, texture, water binding and process tolerance, and which one is chosen decides how far the formula can go.
Soy protein is functionally mature and cost-controllable, but carries its own beany note to mask or use. Pea protein has a lower allergen profile and a more neutral flavour, but different texture behaviour and water binding, so the formula has to be rebalanced. Gluten delivers bite but rules out a gluten-free option.
Beyond flavour and texture, two things travel with the choice: allergen declaration, and the supply stability of that protein source. The first affects the label and which channels can carry it; the second affects long-term cost volatility. Neither is a formulation question, and both constrain the formula.
So the protein source is settled before sampling. Changing it midway is close to starting development again, because texture, water binding and post-kill-step behaviour all have to be revalidated.
Ambient, chilled, frozen: the format trade-off
One formula concept can become three formats, and the choice is not a technical preference. It is a channel and cost decision. The table sets the differences out so the choice has something to rest on.
| Ambient retort | Chilled ready meal | Frozen | |
|---|---|---|---|
| Flavour and texture | Most affected by heat | Best | Depends on freezing rate |
| Shelf life | Longest | Shortest | Long |
| Logistics cost | Lowest | High | Highest |
| Suits e-commerce delivery | ✓ | ✗ | Needs bundling |
| Suits a chilled retail cabinet | ✗ | ✓ | ✗ |
| Suits a commercial kitchen | ✓ | Partly | ✓ |
| Formula constraint | Most | Fewest | Moderate |
| Delivery freshness pressure | Low | High | Low |
What the development stage covers
Development is not only getting the taste right. Each of the items below can become the thing that blocks production later, so they are handled in the same phase.
Umami sources, fat aroma, controlled Maillard development and their ratios
Protein source, water-binding ingredients, texture validated after the kill step
Fully vegan, ovo-lacto or allium-free, with the segregation each implies
Whether the formula survives the target heat curve or freezing condition
Sampling points, storage temperatures, test parameters and acceptance criteria
From measurement or calculation, with allergen and vegetarian class declarations
State the target channel and the vegetarian class first and the development direction is fixed. Those two together decide the format, the kill step and the usable lines, and they are the only conditions no amount of technique gets around.
