Software evolvability
through legacy re-engineering.
Extracting and codifying latent knowledge from complex enterprise systems and using it to evolve them
Research and reverse engineering
The knowledge foundation.
Enterprise systems embody years of decisions about products, processes and operations. That knowledge is distributed across software, data, integrations and operating arrangements.
Research and reverse engineering connect those parts into a supported account of the agreed business domain. Scope follows the process and its consequential dependencies across the estate.
Existing systems can be working well. New capabilities still depend on understanding the rules and dependencies that must be preserved or deliberately changed.
What the work makes explicit
- Business processes, rules and calculations.
- Data meanings, dependencies and interactions.
- Constraints, invariants and service commitments.
- Product and market assumptions, supported by the available evidence.
- Implementation, operating conditions and failure behaviour.
Software evolution
Design and build the target state.
Your enterprise sets the business direction. Living Tensor uses knowledge of the existing estate to propose the technical target state and a practical path to it, with the trade-offs made explicit.
Delivery includes designing and building new software, changing existing systems, integrating both, and bringing the result into operation. Living Tensor carries out the engineering or leads delivery with your team.
The understanding gained through reverse engineering carries into architecture, code and verification: which rules to preserve, which constraints to respect, and which behaviours to change deliberately.
The business continues operating.
Required business behaviour, data and service commitments become explicit design and acceptance criteria.
New software is integrated with the systems that continue to run. The transition accounts for dependencies, data migration, operating responsibilities and recovery if a step fails.
Implementation is verified against those criteria. Any maintenance window or irreversible step is agreed, and specifications are updated to reflect the resulting systems.
What is delivered
Delivery spans the knowledge foundation, the target-state design and working software. Each engagement defines the result and the conditions for completion.
- Codified knowledge
- Processes, rules, constraints and dependencies recorded in documentation, design specifications or executable specifications, with supporting evidence and material uncertainty.
- Target-state design
- A proposed architecture and implementation plan grounded in that knowledge, with the design decisions, trade-offs, integration and transition made explicit.
- Working software
- New and changed software, implemented, integrated and verified, with source code and the materials needed to build, operate and maintain it.
Knowledge can be commissioned on its own. Software delivery carries that foundation through design, implementation and transition into operation.
Method
Reverse engineer. Codify. Evolve.
Agree the result, the relevant domain and the conditions for completion. Evolution work connects knowledge of the current estate to a proposed target state and its implementation.
Reverse engineer
Trace processes, rules and dependencies through code, data, integrations and operating records. Use reliable knowledge already available and resolve gaps relevant to the agreed result.
Codify
Express the findings in documentation, design specifications or executable specifications. Distinguish what the systems do, what the business requires and what remains uncertain.
Propose and build
Use that foundation to propose the target state and transition. Design, implement, integrate and verify the agreed software, updating the specifications as it evolves.
Research produces useful findings throughout the work. Engineering delivers working software, with verification and updated specifications as it progresses. Material uncertainty is made explicit where it affects the result.
Engagement
Useful results early.
A completed job.
Less dependency.
The objective is to complete the agreed work and make continued dependence on Living Tensor unnecessary as early as responsibly possible.
Building a permanent consulting presence is not the business model. Further involvement follows a defined need and your choice.
- Delivery throughout the work
- Research delivers useful findings; implementation delivers working software. Editable sources and explanations are shared throughout, with knowledge transfer during delivery.
- A result your team can take over
- Executable specifications include their source and the means to run them. Software delivery includes source code and the agreed build, operation and transition materials.
- The ability to stop or change provider
- Completed work remains usable. Unfinished work is handed over with its evidence, assumptions and open questions so your team or another provider can continue.
- A defined end
- Completion means the agreed result and coverage have been delivered and responsibilities transferred. A recommendation to buy more work is not a substitute.