Engineering | From engineering
How to estimate AI-native software projects
Estimating under uncertainty: the pitfalls, a five-step method, and what changes with AI
Offerings
AI transformation
Latest
More case studies
Our services
Services
We embed senior engineers, put frontier AI behind them, and ship working software in weeks - then prove it with the numbers.
Learn moreOfferings
AI transformation
Case study An Australian Securities Exchange link, rewired in seven weeks Two forward-deployed engineers, first commit to the accreditation gate in 47 days. Read the case study
Demo See realfast in action Bring us the problem that's urgent and expensive. We'll tell you whether we can move it. Book a meeting Case studies
The work
Real implementations that shipped to production and moved a real metric.
Learn moreLatest
More case studies
Company
The firm
Who we are, who backs us, how we hire, and how we keep your data safe.
Learn moreResources
Writing and talks
What the team writes, and where the founders have been interviewed.
Learn moreResources
Latest post After AI, There Is No Average With AI, the bell curve an enterprise is built on becomes a power law: a few people produce most of the value, and most of the rest produce slop. Read the post
Latest conversation What is an AI-native firm? Subtext by Zerodha, with Sidu Ponnappa. See all conversations
Principal Engineer
2 posts · Engineering · Writing since Apr 2025
Principal Engineer at realfast. Writes about the engineering underneath agents that work in production.
Engineering | From engineering
Estimating under uncertainty: the pitfalls, a five-step method, and what changes with AI
By Prashant Mittal | Oct 3rd 2026 | 17 min read
Engineering | From engineering
Best practices and tricks of the trade that ensure your Agentforce agents operate accurately and consistently, even under complex conditions.
By Prashant Mittal | Apr 14th 2025 | 3 min read