Solana Arbitrage Bot Optimization

I'm building a Solana arbitrage trading bot using the Jupiter API.
The bot can already detect profitable arbitrage opportunities across token pairs.
However, I need an expert to help design and implement the optimal trade sizing strategy.
The challenge is that if I input too much capital into a trade, slippage can erode profits or even cause losses.

What I need help with:
- Designing a method to dynamically calculate the best input size for each arbitrage opportunity.
- Querying Jupiter quotes for different trade sizes to predict slippage.
- Building a simple profit-maximization model (e.g., finding the input size that gives the maximum net profit).
- (Bonus) Optimizing speed and transaction protection using Jito bundles.
- (Bonus) Modeling liquidity curves for better slippage prediction.

Requirements:
- Strong experience with Solana, SPL tokens, and Jupiter aggregator API.
- Understanding of slippage, liquidity pools, and how AMM price impact works.
- Experience with Rust (preferred) or TypeScript.
- (Bonus) Knowledge of Jito bundles and MEV protection strategies.
- (Bonus) Prior experience building or optimizing trading bots on Solana or EVM chains.

Deliverables:
- Discussion and solution design for optimal trade sizing.
- Help me integrate the solution into the bot (can be collaborative; I can do coding if needed).