Skip to content

FBI Sentinel Programme Saved by Agile?

May 31, 2012

FBI Sentinel Programme Saved by Agile?

A story of how an ‘Agile’ approach succeeded in two years, after 10 years of ‘Waterfall’ failure and $800m of waste…

Full case study PDF Brian Wernham – Talk to APM Milton Keynes on 20 June 2013 – The FBI Goes Agile v02.

The story briefly:

Two failed attempts were made to create a new case management system for FBI agents to use in the wake of the Oklahoma City bombings.  The first project ,‘VCF’, was canned in 2005 with a $170m write-off.

A second attempt at building the system, now called ‘Sentinel’, was attempted, again following the waterfall lifecycle described in the FBI’s ‘Life Cycle Management Directive’.  This time even more design would occur up-front with a comprehensive blueprint for the future IT environment and a fixed specification at the beginning with a variable budget.  The committed budget was a staggering $425m, which included $120m for the FBI’s Sentinel Program Office, which would oversee implementation.

Soon after starting, the project started to slip, so the new CIO, Chad Fulgham, who had joined in 2008 tried to gain clarity over the project by requiring the supplier (Lockheed Martin) to deliver in incremental phases.  However the only part of 4 of the planned 18 workflows had gone live with usability, performance, and quality problems.  Finally, in 2010 Fulgham decided to stop the project when estimates showed that Sentinel would cost at least an additional $351m to complete on top of the $405m already spent, and would take another 6 years.  Confidence in the system was so low that the use of the few functions actually delivered was officially optional.  Between July and August 2010 only 132 new forms were generated in the Sentinel system – less than 1% of the 14,831 documents entered into the ACS in the same time period.

In September 2010 Fulgham announced that he would take direct management of the development and use the Scrum method.  He removed all Lockheed Martin personnel from development work on the project and started to directly supervise the sub-contractors.

The number of development staff was reduced from over 125 to just 55, and progress started to be made.

The work was now organized into 670 ‘user stories’ to be developed over 21 sprints – each 2 weeks long.  These user stories were documented in the Sentinel Product Backlog mapped to the original Systems Requirements Specification.  Each story was allocated a number of “story points” based on the difficulty of the work associated with each.  At the start of each sprint the development team identifies which stories will be developed during each sprint.  And at the end of each two week regardless of whether all work is complete, the development team must test and demonstrate the system.  Only those stories that passed those tests were claimed to have been completed.

Delivery has been hampered by a suspension of all system changes around the tenth anniversary of 9/11 last year, and there are still concerns about its performance and availability until the standard five-year refresh of computer hardware is completed.

According to plans Sentinel now should be ready to use, and the proof (or not) of this new agile approach is about to be made public…

About Brian

Brian Wernham is author of “Agile Project Management for Government” to be published by Maitland and Strong on 31st July 2012.  (Pre-order at  or

Subscribe to my blog in one click:



Wernham, Brian (2012): Agile Project Management for Government. Leadership Skills for Implementation of Large-scale Public Sector Projects in Months, Not Years. New York, London: Maitland and Strong

OIG (2002): An Investigation of the Belated Production of Documents in the Oklahoma City Bombing Case. March 19, 2002., updated on 14/03/2002.

OIG (2006a): Sentinel Audit I: The FBI’s Pre-Acquisition Planning for and Controls Over the Sentinel Case Management System. March 2006, Report 06-14., updated on 3/10/2006.

OIG (2006b): Sentinel Audit II: Status of the Federal Bureau of Investigation’s Case Management System. December 2006, Report 07-03., updated on 30/11/2006.

OIG (2007): Sentinel Audit III: Status of the Federal Bureau of Investigation’s Case Management System, Audit Report 07-40. August 2007, Report 07-40., updated on 29/08/2007.

OIG (2008): Sentinel Audit IV: Status of the Federal Bureau of Investigation’s Case Management System. December 2008, Report 09-05., updated on 1/05/2009.

OIG (2009): Sentinel Audit V: Status of the Federal Bureau of Investigation’s Case Management System. November 2009, Report 10-03 Redacted., updated on 11/09/2009.

OIG (2010a): Sentinel Audit VII: Status of the FBI’s Implementation of the Sentinel Project. October 2010, Report 11-01, updated on 15/10/2010.

OIG (2010b): Sentinel Audit VI: Status of the FBI’s Implementation of the Sentinel Project. March 2010, Report 10-22., updated on 30/03/2010.

OIG (2011): Sentinel Audit VIII: Status of the FBI’s Implementation of the Sentinel Project. December 2011, Report 12-08. US OIG., updated on 22/12/2011.

© Brian Wernham 2012 CC BY-NC-ND

From → Agile Governance

  1. Susan Atkinson permalink

    VCF – and now Sentinel – appear to exhibit all the classic symptoms of a waterfall project gone wrong. Over-budget, over-schedule, no product to speak of, and probably a thoroughly demotivated and demoralised team.

    I wonder, however, whether the FBI’s interpretation of Scrum is going to fare much better? A long product roadmap with user stories specified way in advance and lots of emphasis on working practices still misses the point.

    The FBI has a worthy vision – to create a depository for FBI records, allowing agents to share valuable information that might identify links between cases under investigation – but the FBI has got bogged down in the ‘what’, and is losing sight of the ‘why’.

    Agile promotes iterative and incremental development. In a fast-paced and complex environment this approach allows teams to develop new products, and to get feedback, fast. But unless you articulate clearly and focus on the ‘why’, the products are likely to be the wrong products and the feedback will inevitably be vacuous and meaningless.

    I am interested to see the results of the FBI’s change of approach. However, I don’t believe that this will provide proof or otherwise of whether Agile works. It will simply provide evidence of how far along the journey the FBI has travelled in terms of becoming truly Agile …

    • Re: Susan Atkinson: <>

      I agree – the team is using Scrum at a micro-level, with sprints, a product backlog and so on, but at the more strategic level they went a long time 2011-12 without implementing any operational incremenets output by the team. Hopefully we will see some sucess in the near future… My book will cover this in more detail … (

    • It looks like good news from the FBI – they have reported that Sentinel project has delivered after 10 years of waterfall failure, and 2 attempts to implement, it was the third attempt using Agile that delivered in 2 years and at 1/3 of the amount of money thrown away on the first two attempts.


  2. Bud Relihan permalink

    Brian — Just bought your book (should arrive shortly). I hope the FBI is implementing — taking POTENTIALLY customer facing code and making if ACTUALLY customer facing. That’s how Agile really changes the delivery of value…most important features actually put to use BEFORE the end of the project. Please keep us updated. Thanks!

Trackbacks & Pingbacks

  1. How the FBI Proves Agile Works for Government Agencies | Geeklin

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: