Struggling with ASIN targeting scalability for multi-category products using Amazon SP-API

hey everyone,

we're managing Amazon Ads for a portfolio of several brands, each with hundreds of SKUs, and are heavily reliant on dynamic ASIN targeting. it's become pretty clear that to keep up, we're attempting to scale our operations using custom scripts built on the Amazon SP-API. this is for everything from bid adjustments to campaign creation, and frankly, it's a beast.

the core issue is the immense challenge in dynamically updating and managing ASIN targets across thousands of product variations and hundreds of campaigns. our current semi-automated processes are hitting severe bottlenecks, and even our custom SP-API scripts are frequently encountering rate limits or becoming overly complex and brittle to maintain. honestly, it’s a constant battle against the limitations of the Amazon Sponsored Products API, and we're just not seeing the throughput we need.

specific challenges we're facing include:

• efficiently identifying new, high-potential competing ASINs for targeting, often requiring real-time market data that's hard to integrate.
• automating the addition and removal of targeted ASINs at scale without constantly hammering the API.
• maintaining granular control over bid adjustments at the individual ASIN level across a massive product portfolio.
• dealing with Amazon's inherent data lag and attribution complexities, which makes real-time optimization of these targets a nightmare.

so, i’m seriously seeking advice on:

• proven best practices for large-scale ASIN targeting automation, specifically for high-SKU accounts.
• effective strategies or architectural patterns to circumvent SP-API rate limits for frequent, necessarry updates.
• any recommended third-party tools, platforms, or custom architecture designs that have successfully tackled this specific scalability problem.
• experiences or insights into using machine learning models for dynamic ASIN identification, bid management, and overall campaign optimization in this context.

seriously looking for practical, technical solutions to streamline this entire operation. any deep insights would be hugely appreciated. help a brother out please...