Startup Product • 2024Tech + Market AnalysisProduct DesignComputer Vision
Aqua 4.0 (ShrimpVision)
Mobile-first AI product to automate shrimp larvae counting and generate biomass + operational insights.
Role: CTO — led technical strategy, product architecture, model feasibility analysis, and mobile UX concept design.
Counting accuracy target
~95% beta → goal 98%
Hardware requirement
Smartphone camera only
Pilot scope
12 hatcheries + 150–200 farms planned
Scale design
40k+ farms, multi-region roadmap

Problem
Shrimp larvae counting is still manual and inconsistent, leading to stocking errors, feed inefficiency, pricing disputes, and unreliable farm planning. Hatcheries need a fast, repeatable, and accessible way to measure counts and biomass without expensive equipment.
Approach
- Defined the core product workflow: scan larvae → AI counts + sizes → biomass estimate → downloadable report + insights.
- Led technical feasibility analysis for smartphone-only deployment, model constraints, and image quality handling.
- Designed the product architecture including mobile UX flow, reporting logic, and CRM-style tracking of farm history.
- Benchmarked competitors and positioned the product around accessibility and scalability instead of hardware dependency.
- Worked with the team to define the pricing model, pilot rollout plan, and product roadmap.
What I Owned
- Led technical feasibility analysis for smartphone-based counting
- Designed product workflow and mobile UX concept
- Defined system architecture and deployment approach
- Contributed to pricing, rollout, and go-to-market strategy
Impact
- Transforms larvae counting into a repeatable digital workflow instead of manual estimation.
- Extends beyond counting into biomass insights and farm-level decision support.
- Creates a scalable mobile-first product viable for small and mid-size farms.
- Demonstrates end-to-end product thinking: problem framing, technical feasibility, UX design, pricing, and rollout.
Deployment & Monitoring
- MVP architecture: mobile capture → preprocessing → model inference → cloud sync → report generation.
- Pilot rollout planned across hatcheries and farms with structured feedback collection.
- Product feedback loop via in-app signals, field agents, and usage analytics.
- Scale roadmap: India launch → Southeast Asia → Latin America → expansion to additional species.
Risks & Tradeoffs
- Image quality variation (lighting, water clarity, device differences) requires robust preprocessing and model updates.
- Adoption risk if workflow is slower than manual methods — UX simplicity is critical.
- Connectivity constraints require offline-first capability and reliable sync behavior.
Tools
Computer VisionMobile UX designProduct architecture planningGoogle Cloud integrationFlutter & Firebase