The Shift From AI Adoption to AI Optimization

₹78 crore a month. That's what one large Indian enterprise was paying for AI before anyone asked: do we actually need all of this? That number comes from a recent Economic Times piece that paints a picture most CFOs haven't seen yet. Indian firms alone are consuming 400 billion tokens a month. The average developer burns through 51 million. Power users hit 380 million. And the bill is growing faster than anyone's ability to track it. The era of AI adoption is over. The era of AI optimization has begun. Here's what's actually happening on the ground right now: Companies are discovering that 80% of their AI tasks can be handled by smaller, cheaper models. Frontier models are overkill for routine queries. Enterprises are switching to open-source alternatives like DeepSeek and Qwen for low-reasoning tasks. Microsoft cancelled its Claude Code licenses after token costs ran past their annual AI budget months ahead of schedule. Amazon shut down an internal leaderboard called KiroRank after employees gamed it by running pointless AI tasks to boost their rankings, a practice they called "tokenmaxxing." The word showing up in every boardroom is "tokenomics." This isn't companies pulling back on AI. It's companies getting smart about AI. But here's the gap nobody is talking about: all of this optimization is happening inside engineering teams. The engineers are swapping models, adjusting token limits, running benchmarks. And the CFO? The CFO is still getting one invoice per provider with zero breakdown of where the money went, which team spent it, or whether any of it was wasted. Engineering has observability tools. Finance has nothing. That's why we exist. Spend360 gives finance teams a single dashboard across every AI provider, with anomaly detection, budget controls, business unit breakdowns, and per-member spend attribution. We built it for the person who writes the check, not the person who writes the code. The question companies should be asking right now isn't "are we using AI?" It's "are we using the right AI, for the right task, at the right cost, and can finance actually see it?" If your company is spending on multiple AI providers and nobody on the finance side has a single view of what it costs, that's the problem we solve. Hit reply and tell me: how does your company handle AI cost governance today? I read every response. - Srini Founder, Spend360