BitLabs Technologies

We build, integrate and maintain line-of-business software for mid-sized and enterprise organisations.

BitLabs is a software engineering firm based in Nairobi. We work with national agencies, distributors, financial services providers and growing companies across East Africa. We build custom applications, connect systems that were never designed to work together, and support them in production.

What we do

Most engagements combine several of these, and most begin with an assessment of what is already in place before anything new is built.

Automation & AI
Business process automation and applied AI: document processing, data classification, routing and report generation. We automate the high-volume steps in a process and leave judgement and exception handling with your team.
Websites & Online Platforms
Customer portals, self-service platforms and public-facing systems built on React and Next.js, with the accessibility, performance and browser support that public-sector and regulated deployments require.
Mobile Apps
Android and iOS applications from a single React Native codebase. Built for teams working away from an office: continued operation without connectivity, and clean reconciliation once the network returns.
Cloud & Hosting
Deployment, monitoring, backup and disaster recovery on AWS and DigitalOcean, or on in-country infrastructure where data residency is a requirement. Ongoing maintenance under agreed response times.
Bulk SMS
High-volume SMS for notifications, alerts and confirmations, integrated directly into your systems rather than operated as a separate tool. Includes delivery reporting and a per-message audit trail.
System Integrations
Connecting ERP, CRM, payment channels and third-party services into a coherent landscape. REST and event-driven interfaces, with reconciliation and failure handling designed in from the start rather than retrofitted.

Where we are usually brought in

These are the six situations that account for most of our work. Each one expands into what is normally behind it, what it costs while it persists, and how we approach it.

A core system no longer matches how the business operates, and full replacement is not commercially viable.

The system was specified for an earlier operating model and has since been extended past its design. Replacing it outright means a multi-year programme and a cutover risk the business will not accept, so the usual outcome is that nothing happens and the workarounds accumulate.

National government agency · 2024 · pending case volume reduced 60% across two quarters

Cost of the status quo

  • Process knowledge held informally rather than in the system
  • Workarounds in spreadsheets outside any audit trail
  • Change requests quoted in months

How we approach it

  • Establish which functions genuinely depend on the core system
  • Move capability out in stages, each independently testable and reversible
  • Keep the system of record authoritative until the last stage
Discuss this with us
Figures from different systems do not reconcile, and reporting has become manual and contested.

Each system holds a defensible version of the same fact, recorded at a different time under different rules. Without a defined source of truth and a reconciliation process, every reporting cycle turns into a negotiation between departments.

Beverage distributor · 2024 · monthly stock reconciliation reduced from days to same-day

Cost of the status quo

  • Senior time spent reconciling rather than deciding
  • Board and regulatory reporting that cannot be re-derived
  • Disputes that surface at month end with no evidence trail

How we approach it

  • Define the authoritative source for each contested figure
  • Automate reconciliation and record every exception
  • Preserve lineage so any reported number resolves to its source record
Discuss this with us
An integration backlog is holding up initiatives that depend on systems exchanging data.

Integrations were built point-to-point as each need arose. The number of connections now grows faster than the team can maintain them, and every new initiative inherits the full weight of that history before it can start.

Cost of the status quo

  • Delivery dates set by integration work rather than by the initiative
  • Failures that surface downstream, long after the cause
  • Vendor changes that require unplanned remediation

How we approach it

  • Map the existing interfaces and where they actually break
  • Consolidate onto documented, monitored interfaces
  • Build reconciliation and retry handling into each one
Discuss this with us
A vendor-built system can no longer be extended, and the original supplier is no longer engaged.

The handover covered access but not knowledge. Without documentation, test coverage or a working local environment, any change carries unquantified risk, so the system is frozen in place while the business moves on around it.

Cost of the status quo

  • A system nobody is willing to change
  • Commercial exposure to a single supplier
  • Security patching deferred because the impact is unknown

How we approach it

  • Recover the build, establish a working environment and document what exists
  • Add test coverage around the areas that will change first
  • Resume delivery once change can be made safely
Discuss this with us
Processes that hold at today's volume will not survive growth, acquisition or a new mandate.

Manual steps that are tolerable at current throughput become the binding constraint once volume rises. The failure is rarely gradual: it appears as a queue that stops clearing, usually at the least convenient moment.

Beverage distributor · 2024 · keg movements tracked across a national distribution network

Cost of the status quo

  • Headcount scaling in step with volume
  • Service levels that degrade under peak load
  • Integration or expansion plans constrained by operational capacity

How we approach it

  • Instrument the process and identify where it will bind first
  • Automate the steps carrying the volume, leave judgement with people
  • Load-test against projected rather than current throughput
Discuss this with us
Audit, regulatory or donor requirements cannot be evidenced from the systems currently in place.

The systems record outcomes but not the decisions behind them. When an auditor asks who approved something, on what basis and when, the answer has to be reconstructed from email and memory.

National government agency · 2024 · full chain of custody per case record

Cost of the status quo

  • Audit findings that recur each cycle
  • Evidence packs assembled by hand under deadline
  • Approvals that cannot be demonstrated after the fact

How we approach it

  • Record decisions and approvals as first-class data, not as side effects
  • Make the audit trail immutable and queryable
  • Produce evidence packs directly from the system
Discuss this with us

Results, with their working shown

Each figure below expands to show the baseline it was measured against, the method used, and the factors we cannot rule out as alternative explanations. Client references are available on request and under NDA where the engagement requires it.

Case backlog reduced 60% in two quartersopen
Client
National government agency (named on request)
Baseline
1,431 open cases · Mar 2024
After
572 open cases · Sep 2024
Measured by
Counted from the agency's own case registry rather than from our system, at the start and end of the period.
Confounders
The agency also increased case-handling headcount in July 2024. The two effects cannot be isolated from the available data, so this figure should be read as the combined result of both.
Verify
Client reference letter available on request

Sectors we work in

Government and public sector

Case management, licensing and workflow systems for national agencies and county governments, built to hold up under audit. We are registered on eGP; our statutory documents are published on the company profile rather than held behind a request form.

Discuss a project in this sector

Distribution, logistics and manufacturing

Asset tracking, stock reconciliation and field operations across distributed networks. Immutable movement records and chain of custody, so balances can be evidenced rather than negotiated at period end.

Discuss a project in this sector

Financial services and fintech

Payment channel integration and automated reconciliation against a single ledger, with each transaction traceable to its source channel. Built for the controls and audit requirements that regulated operations carry.

Discuss a project in this sector

NGOs and donor-funded programmes

Field data collection that continues to operate without connectivity, and reporting produced in the donor's required format. The underlying record is preserved so any figure in a report can be traced to source during evaluation or audit.

Discuss a project in this sector

Discuss a project with us

Send us an outline of what you are trying to achieve and the constraints you are working within. We will respond with an assessment of whether we are a suitable fit, and where we are not, what we would recommend instead.

Alternatively, use the contact form to send a written brief.