Part 3 - From SDLC to BDLC

Agentic AI impacts on the Software Development Life Cycle

In this article, Part 3 of a 4-part series, we’ll focus on the Process and SDLC impacts of Agentic AI:

• How the SDLC will be replaced by the BDLC (Business Delivery Life Cycle)
• What core executive mindset changes will be needed
• How will the roles change
• What about funding and metrics

Part 1 examined what the C-suite is truly seeking, the distinction between hype and reality regarding Agentic AI, and why code speed is not the bottleneck. Part 2 discussed the People and Workforce design aspects.

Software changes that took ~6 weeks to complete can now be done in ~6 hours. That’s 168 times (16,700% ) faster.

10x faster was a moonshot earlier this year, now even 100x is not ambitious enough!

Current SDLC process

 In most organizations, including Agile ones, the SDLC looks something like this, with slight variations:  

Characterized by:

• Long lead times in months, with most of the time spent outside of the ‘develop’ phase.
• Large, costly year-long programs of work and large backlogs of small but important changes
• Numerous siloed roles; 12-15 at least. Every skill or task has become a separate role
• High costs and a large workforce; people cost approximately 80% of the total IT spend, with many enablers
• Complex frameworks, difficult prioritization models, and overloaded teams
• Almost no customer or end user contact by the delivery teams till software release
• Business value projections are rarely met as the estimates are not based on real customer validation, feedback, and early tests

If coding can now be done 15,000 %+ faster, the processes above and below the coding step would also need to be sped up significantly, if we are to reap the benefits of this coding speed improvement.

From SDLC to BDLC

With Agentic AI proving itself to be more than just a fad, the future will look much different.

 IT will be the business. The SDLC will be replaced by the BDLC.

While I have linearly represented this, it would in fact be a continuous loop.

 So, what would need to change?

Budget reallocation

Budgets for 2026 are being agreed as we read. For most companies, there are no big budget increases for the use of AI.

Most CFOs want the benefits of AI but are not willing to put up the funding for it. The adage ‘show me your budget and I’ll tell you your real strategy’ still applies.

What this means is that CIOs will need to find savings to invest in AI. As normally 70-80% of the IT budget is people costs, the people reduction has started.

We must not forget that AI agents and the use of LLMs behind them will cost almost as much as people. Companies like AWS, GCP, and Microsoft will provide agentic services on call, including secure LLM access, at a fee.

Here is what the change in budget could look like:

Budgets would remain cost-neutral, with a significant spend profile shift away from people and towards AI.

Conclusion

With Agentic AI enabling us to deliver 1,500% faster, we will need to change our entire approach to software development to take advantage of it.

Our mindset will need to shift from big programs, defined and estimated in detail upfront, to a faster, more effective experimental approach.

Business and IT would need to work closely together, and there would be a need for one way of working across Business and IT. A Business Development Life Cycle rather than a Software Development Life Cycle.

The number of roles would reduce as AI Agents take over much of the design, coding, and testing work. Leaders would need AI awareness and technical knowledge to make decisions on risk and ROI.

The well-known saying ‘AI will not take your job, but people with AI knowledge will’, should motivate us to AI-upskill and get more technical where necessary.

On a more positive note, the bar has lowered for disruptive new business opportunities, and AI-enabled entrepreneurship would be a great alternative for many. Look for a problem area where people and organizations are slow to change, and disrupt these slow-moving incumbents with innovative AI.

Funding and governance would need to keep up in speed, adaptability, and responsiveness to enable faster time to market.

Organizations that change their overall way of working will be the winners.

In part 4, I’ll cover the technical changes needed to deliver 100x faster.