Ø5Cases

Measured results in operations that cannot stop.

Ten projects in mobility, credit, retail and fitness. Each with the problem the operation had, what Blite did and the number that changed.

80%

less time to open a rental contract

99,95%

availability after the migration to AKS

95%+

of fleet traffic fines assigned within the legal deadline

1month

to build a complete technology team

Filter by service
  1. Ø01Mobility · Vehicle rentalContracts opened in 2 minutes at the counter of more than 160 stores.80%reduction in contract opening time
  2. Ø02Financial services · Vehicle financingFrom no app to four in the stores, with Open Finance built into the architecture.4apps published on the App Store and Google Play
  3. Ø03Credit · Large enterpriseThree credit operations with the same standard and analysis in 2 hours.75%less analysis time, from 8 to 2 hours
  4. Ø04Mobility · National fleetFleet traffic fines read by AI and assigned within the legal deadline.90%+of fines processed without human intervention
  5. Ø05Mobility · Customer serviceWhatsApp as the main digital channel, with more leads and less waiting.52,1%of customer contacts moved to WhatsApp
  6. Ø06Mobility · Digital operation30 squads, 23 business areas and a single view of the operation.70%+more delivery predictability
  7. Ø07Fitness · National gym chain40 microservices migrated to AKS without stopping the chain.99,95%availability in critical services
  8. Ø08Mobility · Software engineeringClaude in the development cycle, with guardrails and auditing.40%more items delivered per squad
  9. Ø09Retail · Omnichannel operationSix critical roles stalled for 90 days. Full team in 1 month.60%less average time to close critical roles
  10. Ø10Mobility · LM · SAP S/4HANAThe team that migrated S/4HANA is the same one running the operation today.7dedicated professionals

10 cases

Ø01Mobility · Vehicle rental

Contracts opened in 2 minutes at the counter of more than 160 stores.

One of Brazil's largest car rental companies replaced its store front core system without interrupting service nationwide.

  • Tech Consulting
  • Product Delivery
80%reduction in contract opening time
5.150contracts on a peak day
160+stores running on the new core
30%higher customer satisfaction

Context

Vehicle pickup depended on the system that opens the contract, invoices and validates payment. As volume grew, every minute of slowness at the counter turned into a queue.

Challenge

  • Opening a contract took about 10 minutes
  • Visible queues at peak hours
  • A platform with no headroom for demand swings
  • Any instability hit hundreds of stores at once

The switch had to happen with the national operation running.

What Blite did

  1. 01Technical diagnosis of legacy bottlenecks and architecture, performance and integration guidelines
  2. 02Dedicated teams building the new core in short cycles
  3. 03Integration with invoicing, financial validations and business areas
  4. 04A transition with no downtime, with stability tracked store by store
Time at the counterOpening one contract
Before
10:00
After
2:00
3.000contracts per day, on average
5.150contracts on the peak day
160+stores on the new system

Before and after

Before
  • Legacy system at its performance limit
  • Contract in about 10 minutes
  • Counter queues at peak hours
After
  • A platform ready to scale
  • Contract in about 2 minutes
  • An average of 3,000 contracts per day, with stability

Ø02Financial services · Vehicle financing

From no app to four in the stores, with Open Finance built into the architecture.

A finance company with two brands gained a complete mobile channel for individual and business customers, within Central Bank requirements.

  • Product Delivery
4apps published on the App Store and Google Play
1Flutter codebase for both brands
100%of the Open Finance consent flow
0internal mobile team required

Context

The entire digital relationship with customers ran through the web portal: managing financing contracts, payments, prepayments and personal data updates.

Challenge

  • Go from zero to four apps: two brands, iOS and Android
  • Independent visual identities, without doubling maintenance cost
  • Open Finance required by the Central Bank
  • PCI DSS controls for payment data
  • No internal team set up for mobile

Blite owned the whole product, from discovery to store publication.

What Blite did

  1. 01Java BFF orchestrating authentication, notification and customer portal microservices
  2. 02Integration with Salesforce and external APIs through an API Gateway
  3. 03Complete Open Finance consent flow
  4. 04Encryption in transit and at rest, and traceability of every transaction
  5. 05CI/CD with E2E tests and Dynatrace monitoring on backend and frontend
One base, four appsFlutter
Single codebase
Brand AiOSAndroid
Brand BiOSAndroid
Open Finance consentFlow required by the Central Bank
  1. 1Institution selection
  2. 2Authentication
  3. 3Authorization
  4. 4Consent recorded
Java BFFMicroservicesAPI GatewaySalesforcePCI DSSDynatraceE2E tests

Before and after

Before
  • Customer relationship only through the web portal
  • Payments and contracts depended on service staff
  • No mobile presence
After
  • Four live apps for two brands
  • Customers manage contracts and payments on their phones
  • One team maintaining both brands on the same base

Ø03Credit · Large enterprise

Three credit operations with the same standard and analysis in 2 hours.

Three independent credit areas now decide with the same criteria, approval levels and traceability.

  • Product Delivery
  • Squads & Governance
75%less analysis time, from 8 to 2 hours
40%of approvals with no manual intervention
2.151analyses processed per month
80%less exposure from uneven criteria

Context

Each vertical had its own methodology, controls and team for individual and business credit. As the company grew, approvals became inconsistent and leadership had no aggregate view of the process.

Challenge

  • Three independent credit methodologies
  • Approvals with no formal approval levels
  • Fraud risk with no control layer
  • Analysis taking 8 hours
  • Zero centralized visibility for leadership

Unify without erasing what is specific to each area.

What Blite did

  1. 01Credit criteria unified across the three verticals, with a common language
  2. 02Credit engine with rules for individuals and businesses and decisions in seconds for eligible cases
  3. 03Facial biometrics and anti-fraud in the flow, with checks against credit bureaus and the Federal Revenue Service
  4. 04Analysis desk with a structured form for cases outside the automated path
  5. 05Approval levels by amount and a real-time panel with SLA visible to leadership
Unified decision flowIllustrative
Vertical 1Vertical 2Vertical 3
Credit engineCommon rules for individuals and businesses · credit bureaus and Federal Revenue · biometrics and anti-fraud
40%automatic approval
60%analysis desk with a structured form
Approval level by amountDocumented, traceable decision
Analysis time
8 h
2 h

Before and after

Before
  • Three areas with their own rules
  • Analysis in 8 hours
  • Decisions with no record or approval level
After
  • One platform for the three verticals
  • Analysis in 2 hours, 40% automated
  • Every decision documented and traceable

Ø04Mobility · National fleet

Fleet traffic fines read by AI and assigned within the legal deadline.

One of Brazil's largest car rental companies now treats traffic fines as managed data, with Claude reading the documents.

  • Data & AI
  • Squads & Governance
90%+of fines processed without human intervention
95%+assigned within the legal deadline, previously below 70%
95%+AI classification accuracy
3months to automate the process

Context

Brazil has no single agency that consolidates fleet traffic fines. Each notice arrives with a different format, deadline and channel, and a fine not assigned in time doubles by law. The exposure exceeded tens of millions of reais a year.

Challenge

  • Thousands of fines a month from municipal, state and federal agencies
  • Documents with no common standard across agencies
  • Reading, classification and driver assignment done by hand
  • Short legal deadlines
  • No management view of repeat offenses or risk by branch

Absorb the volume, normalize the formats and automate the cycle from receipt to assignment.

What Blite did

  1. 01Continuous ingestion with Kafka and processing in Databricks
  2. 02OCR for reading and Claude for classification and extraction of type, place, date, deadline and driver
  3. 03Identification of the actual driver from fleet data
  4. 04Critical deadline alerts and risk-based prioritization
  5. 05Executive dashboard of patterns, repeat offenses and risk by branch and driver
From notice to assignmentPipeline
  1. Agenciesmunicipal, state and federal
  2. Kafkacontinuous ingestion
  3. Databricksprocessing and analytics base
  4. OCR + Claudereading and classification
  5. Driveridentified from the fleet
  6. Assignmentwithin the legal deadline
Traffic violation notice
Type
Speeding
Place
Highway, km 112
Deadline
15 days
Driver
Identified
Assigned within the legal deadline
Before · below 70%After · above 95%

Before and after

Before
  • Manual, reactive management
  • Each document handled case by case
  • Less than 70% assigned on time
After
  • Automated cycle with Kafka, Databricks, OCR and Claude
  • Driver identified in hours
  • More than 95% assigned on time

Ø05Mobility · Customer service

WhatsApp as the main digital channel, with more leads and less waiting.

A large mobility company now serves, qualifies and sells through WhatsApp, with real-time data from internal systems.

  • Tech Consulting
  • Product Delivery
52,1%of customer contacts moved to WhatsApp
60%+more leads generated
25%fewer handoffs to human agents
95%less lead waiting time

Context

Traditional service was expensive, slow to respond and relied on people for simple questions. Sales could not predict the volume of opportunities.

Challenge

  • High cost per contact
  • Long response times
  • Reliance on human agents for basic questions
  • Low sales predictability

Treat the channel as a product, with indicators and continuous evolution.

What Blite did

  1. 01Mapping of customer paths and result-oriented conversation flows
  2. 02Integration with internal systems to answer and qualify with real-time data
  3. 03Indicators, rituals and quality criteria for the channel
  4. 04Continuous iteration on adoption, conversion and efficiency after go-live
Service in the channelIllustrative conversation

I need a car for Friday, in Belo Horizonte.

For how many days? I can book it at the store closest to you.

Three days.

I found three options available at the Savassi store. Shall I book the first one?

I'd like to talk about a corporate contract.

Transferred to a consultant with the conversation history

Before and after

Before
  • Simple questions waited in the human queue
  • Leads waiting for an answer
  • Channel with no indicators
After
  • 52.1% of contacts on WhatsApp
  • 30% more qualified leads
  • Sales response 75% faster

Ø06Mobility · Digital operation

30 squads, 23 business areas and a single view of the operation.

One of Brazil's largest mobility companies standardized its agile model and now manages the entire portfolio through Órbita.

  • Squads & Governance
  • Data & AI
70%+more delivery predictability
100%of squads on the standard model
600+people on the new backlog model
20%fewer operational tasks for Product Owners

Context

Each squad had its own process, rituals and definition of done. The board could not compare teams, coordinate dependencies or read the operation without manual consolidation.

Challenge

  • More than 30 squads with no common process
  • More than 23 business areas making IT requests at the same time
  • Inconsistent flow from the board down to the story
  • Metrics that could not be compared
  • Status assembled by hand in spreadsheets

Standardize without losing delivery speed.

What Blite did

  1. 01Corporate agile process templates applied in every squad
  2. 02Standard flow from the executive board to epics, features and technical breakdowns
  3. 03Rollout and training across more than 30 squads and 600 people
  4. 04Órbita integrated with Azure DevOps for performance, maturity and cost in one place
One flow from the board to the storyStandard model
  1. Executive board23 business areas
  2. Epicspriority by value and cost
  3. Featuresportfolio with % complete
  4. Stories and tech stories30 squads, 600 people
Reads Azure DevOps directly
CTO360 view and cost against budget
ManagerPerformance and maturity across squads
Tech leadBurndown, CFD and sprint blockers
See Órbita →

Before and after

Before
  • Squads with their own processes
  • Incomparable metrics
  • Management by manual consolidation
After
  • Common agile model in 100% of squads
  • Whole portfolio read through Órbita
  • Cost and ROI management by workstream

Ø07Fitness · National gym chain

40 microservices migrated to AKS without stopping the chain.

A national gym chain gained a standardized, observable cloud platform with autonomy for its squads.

  • Platform & Cloud
  • Data & AI
99,95%availability in critical services
4xmore frequent deploys
65%less mean time to respond to incidents
25%lower cloud cost

Context

More than 40 services ran on Azure Web Apps with different configurations, inconsistent environments and decentralized deploys. The model limited reliability and product speed.

Challenge

  • Scattered services with heterogeneous configurations
  • Fragmented monitoring
  • CI/CD with no standard and no quality or security gates
  • New microservices depended on centralized infrastructure

Migrate everything in 6 months with the chain running.

What Blite did

  1. 01Managed AKS cluster with namespaces per squad
  2. 02Istio service mesh with distributed tracing across services
  3. 03Corporate observability in Grafana with metrics, logs and traces
  4. 04SonarQube and Veracode as automatic gates in every pipeline
  5. 05Standard instrumentation in 100% of migrated services
40+ services, one platform
Azure Web Apps with different configurationsAKS · namespaces per squad · Istio
squad 1squad 2squad 3squad 4
PipelineBuildSonarQubeVeracodeDeployGrafana

Before and after

Before
  • 40+ services scattered across Web Apps
  • Manual deploys with no gates
  • New service in days or weeks
After
  • AKS platform with Istio and Grafana
  • Deploys 70% faster, with gates
  • New service in hours

Ø08Mobility · Software engineering

Claude in the development cycle, with guardrails and auditing.

Six squads working on critical products adopted a corporate AI platform for engineering in about 3 months.

  • Tech Consulting
  • Data & AI
40%more items delivered per squad
60%less rework from incomplete specs
60%less code review time
20p.p.more unit test coverage

Context

Even with mature DevOps, throughput was held back by incomplete specs, waiting on code review and manual environment setup to validate each PR. Leadership wanted AI in the cycle, with governance.

Challenge

  • Rework from ambiguous specs
  • Code review as a waiting point
  • Uneven test coverage across projects
  • Validation environments created by hand
  • Risk of AI with no guardrails or auditing

AI as a corporate practice, with the same rule for every squad.

What Blite did

  1. 01Spec Driven Development with Claude helping write and refine the specification
  2. 02Unit tests generated and validated on every PR
  3. 03Code review with contextual analysis of the codebase and automatic PR description
  4. 04Ephemeral environment per PR provisioned in minutes
  5. 05Guardrails, usage auditing and integration with repository, CI/CD and issue tracker
AI-assisted cyclePlatform built on Claude
  1. Ø1SpecClaude helps write and refine it
  2. Ø2Codeimplementation from the spec
  3. Ø3PRunit tests generated and validated
  4. Ø4Reviewcontextual analysis of the codebase
  5. Ø5Environmentephemeral per PR, in minutes
  6. Ø6Mergewith automatic gates
Guardrails and auditingQuality, security, scope adherence and traceability of every decision

Before and after

Before
  • Incomplete specs causing rework
  • Code review as a bottleneck
  • AI used without governance
After
  • Spec as the central artifact of the flow
  • Automatic review and tests on the PR
  • More than 80% of squads with the platform in their flow

Ø09Retail · Omnichannel operation

Six critical roles stalled for 90 days. Full team in 1 month.

A large retailer strengthened its technology team with technical rigor and tracking through Telescope.

  • Consultative Headhunting
60%less average time to close critical roles
3daysto the shortlist for each role
88%average fit to the scorecard
100%of hires retained through the guarantee period

Context

Demand for deliveries grew faster than the ability to hire. Roles depended on manual screening, a generalist HR team and managers who were already stretched.

Challenge

  • 6 technical roles open for more than 90 days
  • Poorly matched screening left to the manager
  • Difficulty translating the need into a profile
  • Previous hires that did not stay on the team
  • No visibility into bottlenecks or risk of not closing

Hire fast without giving up technical and behavioral criteria.

What Blite did

  1. 01Role diagnosis: context, seniority, scope and fit criteria
  2. 02Scorecard and competency interview defined before hunting
  3. 03Active search, technical and behavioral checks and background check
  4. 04Shortlist with a consultative opinion and decision support
  5. 05Support from offer to onboarding, all visible in Telescope
Contract timelineCalendar days
Day 0Diagnosis of the 6 roles
3 daysShortlist per role
2 weeks1st hire
1 monthFull team
  1. Diagnosis
  2. Scorecard
  3. Hunting
  4. Short list
  5. Onboarding
Role, candidates, opinion and history in one linkSee Telescope →

Before and after

Before
  • Roles stalled for more than 90 days
  • Decisions with no scorecard
  • Zero visibility into the process
After
  • Shortlist within 3 business days
  • 1st hire in 2 weeks
  • Real-time progress through one link

Ø10Mobility · LM · SAP S/4HANA

The team that migrated S/4HANA is the same one running the operation today.

LM Mobilidade gained a dedicated SAP competence center, built in waves and running continuously for more than a year and a half.

  • Consultative Headhunting
  • Squads & Governance
7dedicated professionals
3ABAP developers
4functional consultants
1,5yearof continuous operation with the same team

Context

LM operates nationwide with SAP at the center of its critical processes. It had to keep the production modules running and migrate to S/4HANA at the same time, competing with large integrators for talent.

Challenge

  • Production modules that cannot stop
  • S/4HANA project in parallel
  • SAP market heated by the migration cycle
  • The right profiles not actively looking

Build the right team and keep the knowledge inside it.

What Blite did

  1. 01Market mapping and outreach to professionals who were not looking
  2. 02Validation of ABAP, functional cases per module and migration experience
  3. 03Comparable criteria so leadership decides on competence
  4. 04Squad formed in waves, with rituals and continuous visibility
  5. 05The implementation team stayed on for operations, with no handover
Competence centerDedicated squad
3 ABAP developers4 functional consultants
FIMMSDin production
  1. S/4HANA projectMigration without stopping the modules
  2. Continuous operationSame team, for 1.5 years
  3. Next phaseBudget allocations, treasury, FX and interest rate risk

Before and after

Before
  • Project and operations competing for the same people
  • SAP profiles scarce in the market
  • Risk of losing knowledge at go-live
After
  • S/4HANA delivered with FI, MM and SD in production
  • Same team from design to operation
  • Next phase starts with the team already inside the environment

Talk to Blite

Does your operation have a challenge similar to one of these?

Talk to a specialist →Back to the index