Jul 2020–Nov 2020
Inventory Specialist
Pier 115 Bar and Grill
Managed inventory and operational tracking, built a custom inventory tool, and supported social media advertising and online visibility.
Joshua Valle
I turn messy operational problems into usable systems
My work spans AI implementation, workflow automation, and full-stack development, from mapping the problem and defining requirements through testing, rollout, and support
- 01 AI implementation From requirements to a working release Discovery · QA · rollout
- 02 Workflow automation Web systems and connected workflows Web apps · APIs · integrations
- 03 Vehicle telemetry Vehicle data, from CAN bus to dashboard C++ · Carloop · InfluxDB · Grafana
- 04 Computer vision Image-processing and model studies Python · OpenCV · model testing
Move the pointer · click empty space for a pulse
2019–2026
Experience
Six roles across technical support, data, development, and AI delivery.
Arcadia University
Community-Based Work-Study
Aug 2019–May 2023
Supported campus and community-based work-study responsibilities while completing a computer science degree.
Work included
- Campus and community support
- Task ownership and follow-through
- Communication across academic environments
- Organized work across multiple stakeholders
What I did
- Built early professional experience in reliability, communication, and task ownership across academic and community environments.
- Balanced work-study responsibilities with a computer science program over four academic years.
- Developed the follow-through, clear communication, and organization that now carry into technical projects.
Hasbro
3D Modeling Intern
Dec 2021–May 2022
Supported modular 3D modeling in a production-style student program using Blender and CAD workflows.
Work included
- Modular 3D modeling
- Blender and CAD workflows
- Blueprint and infrastructure corrections
- Design feedback and iteration
What I did
- Broke complex mold structures into smaller modular model components.
- Corrected blueprint and infrastructure details as the models evolved.
- Worked through visual accuracy, technical constraints, and design feedback in an iterative production-style environment.
Tapestry
Data Analyst Specialist
Mar 2024–Aug 2024
Supported data management, reporting, and operational documentation by maintaining structured records, correspondence, and business data used across internal workflows.
Work included
- Operational reporting
- Data cleaning and validation
- Structured records and correspondence
- Excel, SQL, and Python/Pandas
What I did
- Performed data validation, cleansing, quality assurance, and record maintenance to improve reporting accuracy and database integrity.
- Compiled and processed information for reporting while identifying inconsistencies and keeping business records current.
- Helped improve data-management workflows and documentation practices as operational requirements grew.
DeepShot AI
ML Development Contributor
Aug 2024–May 2026
Supported machine-learning development through model-training workflows, edge-case testing, dataset validation, preprocessing improvements, and AI feature evaluation.
Work included
- Training and testing data review
- Preprocessing and input/output checks
- Edge-case testing
- Feature and model-behavior review
What I did
- Prepared and validated training and testing data, reviewed model outputs, and identified edge cases that affected feature reliability.
- Supported data preprocessing, extraction, and input/output improvements across model-evaluation workflows.
- Partnered with engineering to validate feature behavior, improve dataset quality, and troubleshoot model issues.
- Contributed QA findings that supported continuous refinement throughout the development lifecycle.
Business Solutions Agency
Co-Founder / Lead Developer
Oct 2024–Present
Co-founded and led development for a digital solutions agency delivering websites, custom software, workflow automation, and operational systems for small-business clients.
Work included
- Websites, portals, and e-commerce
- CRM and CMS integrations
- APIs, webhooks, and intake flows
- Reporting, migration, and follow-up workflows
- Testing, launch, and iteration
What I did
- Partnered directly with business owners to analyze workflows, uncover operational friction, and design maintainable technical solutions.
- Built full-stack systems across WordPress, GHL, PHP, JavaScript, Python, AWS, Cloudways, databases, and third-party integrations.
- Connected websites, CRM and CMS platforms, intake flows, reporting, payments, APIs, webhooks, and AI-enabled support experiences.
- Built automations connecting CRM systems, invoicing, order management, customer support, and external services.
- Managed engagements from discovery and requirements through implementation, testing, deployment, client training, launch support, and post-launch improvements.
AJAIA
Technical Project Manager, AI Implementation & Delivery
Apr 2026–Present
Supported AI implementation, product rollout, and workflow automation across client-facing and internal projects.
Work included
- Discovery, requirements, and workflow mapping
- QA, deployment planning, and launch readiness
- Onboarding, access, and support workflows
- Implementation documentation and rollout playbooks
- Stakeholder communication and blocker resolution
What I did
- Worked across stakeholders, product, operations, and engineering to translate workflows, feedback, support issues, and implementation ambiguity into requirements, rollout plans, QA checks, and source-backed project updates.
- Coordinated implementation work across discovery, requirements, workflow mapping, solution design, testing, deployment planning, launch readiness, and post-deployment support.
- Supported onboarding, access provisioning, SSO planning, pilot coordination, usage visibility, monitoring, support paths, and rollout readiness.
- Built onboarding guides, implementation documentation, technical follow-ups, project trackers, meeting archives, knowledge-base content, and scalable rollout playbooks.
- Partnered with clients, executives, operators, product managers, and engineers to prioritize work, resolve blockers, and improve deployment quality.
All experience
Roles, dates, and organizations at a glance.
Jul 2016–Mar 2019
Negotiator & Programmer
Genesis ATM
Supported ATM-related technical operations and transaction workflows, built an automated fund-tracking program, and assisted with product sales and issue resolution.
May 2015–Feb 2016
Social Media Ambassador
Yankee Pizzeria Inc.
Supported online ordering, basic order-management tooling, social media advertising, and customer-facing digital activity.
Selected work
From requirements to rollout
I help move implementation work from an unclear operational problem to a tested release, keeping requirements, owners, blockers, and evidence clear along the way.
Understand the work before prescribing the software
I trace who is involved, which systems are already in the path, where the handoff breaks, and which constraints cannot be ignored. The goal is a problem statement narrow enough to act on.
- Current pathPeople, systems, source material, and handoffs as they exist today.
- ConstraintsAccess, policy, timing, ownership, and technical limits surfaced early.
- Decision recordConfirmed facts separated from assumptions and open questions.
Make the expected behavior testable
I turn the problem into explicit requirements, acceptance conditions, dependencies, and owners. That gives the build team a shared definition of done and gives stakeholders something concrete to review.
- RequirementsUser paths, system behavior, edge cases, and non-negotiable constraints.
- Acceptance checksObservable proof for what must work before a feature is considered ready.
- Build slicesSmall, ordered increments with unresolved decisions kept visible.
Keep technical work tied to the original need
I maintain the working source of truth while implementation moves: what changed, what is blocked, which dependency matters next, and whether the latest build still matches the agreed requirement.
- Implementation trackingVerified status, owners, dependencies, and the next executable action.
- Blocker resolutionAccess, data, environment, product, and coordination issues surfaced early.
- Cross-functional handoffTechnical detail translated without losing the decision behind it.
Exercise the complete journey, not just the happy path
I test the visible workflow alongside access, onboarding, failure states, support readiness, and the fallback path. Gaps remain open until there is direct evidence they are resolved.
- End-to-end QAReal click paths, inputs, outputs, state changes, and repeatable issue evidence.
- Operational readinessAccounts, permissions, onboarding steps, support ownership, and escalation.
- Go / no-go viewConfirmed live behavior kept separate from target state and remaining risk.
Launch with a usable handoff and an honest next step
I prepare the release, onboarding, support, and feedback material around the product so the people taking over know what is live, how to use it, where to report a problem, and who owns what follows.
- Release materialWhat changed, who it affects, and the exact path people should follow.
- Feedback loopClear intake, triage, ownership, and a route back into implementation.
- HandoffOpen items, owners, deadlines, support paths, and next decisions.
Inside the delivery work
This view follows the step selected above
- The question
- What is happening now, where does the handoff break, and which constraints are real?
- What I kept visible
- People, systems, source material, access limits, confirmed facts, assumptions, and open questions.
- What moved forward
- A narrow problem statement, the right owners, and a clear next investigation or decision.
- The question
- What behavior must work, including edge cases, and what will count as proof?
- What I kept visible
- User paths, dependencies, owners, non-negotiable constraints, and observable acceptance checks.
- What moved forward
- Review-ready requirements and ordered build slices with unresolved decisions still called out.
- The question
- Does the latest build still solve the agreed problem, and what is blocking the next usable slice?
- What I kept visible
- Verified status, repository and board evidence, dependencies, blockers, owners, and changed decisions.
- What moved forward
- Blockers were routed early and technical handoffs stayed tied to the original requirement.
- The question
- Can a real user complete the full journey, recover from failure, and get support?
- What I kept visible
- Accounts, permissions, onboarding, failure states, issue evidence, support ownership, and fallback paths.
- What moved forward
- A direct go or no-go view that separated confirmed behavior from remaining risk.
- The question
- What is live, how should people use it, and who owns the next issue or decision?
- What I kept visible
- Release notes, onboarding paths, support routes, feedback intake, open items, owners, and deadlines.
- What moved forward
- A usable handoff and feedback loop that could route real findings back into implementation.
Web and workflow automation for small businesses
Through Business Solutions Agency, I build websites and connected workflows for small businesses, joining the customer-facing experience to the operational steps behind it.
How I approach the work
- 01Map the work
Trace the existing path, the owner at each handoff, and the friction worth solving.
- 02Choose the connection
Use the smallest dependable mix of native tools, APIs, webhooks, and custom code.
- 03Test the full path
Check validation, notifications, failure states, ownership, and what happens next.
- 04Launch and support
Document the workflow, watch the first real use, and improve what the evidence shows.
More about the work
- The need
- Lean client projects where the public website and the work behind it needed to function as one maintainable system.
- What I owned
- Co-founder and lead developer. I handle discovery, scoping, implementation, testing, launch, and iteration with business owners.
- How I built it
- Websites and intake paths connected to the CMS, CRM, API, webhook, reporting, or support tools the workflow actually needs.
- What it produced
- Built maintainable paths between public touchpoints and the operational tools behind them.
What I handled
- Public sites and landing pages connected to intake, CRM, scheduling, email, payment, or support tools.
- Automations for lead capture, outreach, order processing, reporting, data movement, and internal handoffs.
- API, webhook, and native-integration paths selected for maintainability rather than novelty.
- Validation, failure paths, notifications, ownership, and recovery checked before launch.
- Plain-language workflow documentation, launch support, and post-launch iteration.
- APIs
- Webhooks
- CMS/CRM workflows
- JavaScript
- Python
Vehicle telemetry over CAN bus
01 / 05
Actual setup
The real laptop, cabling, and OBD hardware used inside the Lexus
Swipe to see all views
More about the work
- The need
- Arcadia University capstone investigating a web-based path for collecting and analyzing vehicle electrical-system data.
- What I owned
- I developed the Particle C++ firmware, handled OBD-II request and response processing, connected the cloud data path, and built the dashboard presentation.
- How I built it
- 2016 Lexus / Vehicle HS CAN → OBD-II and Carloop → local Wi-Fi → Particle Cloud integration → InfluxDB on a cloud VM → Grafana.
- What it produced
- Collected OBD-II readings such as RPM, vehicle speed, engine load, fuel-level input, and coolant temperature. Recovered artifacts visibly show RPM, speed, and engine load in Grafana during a recorded drive, while source and console materials document the additional readings.
What I handled
- Requested standard OBD-II PIDs and processed ECU responses in C++.
- Sent readings through Carloop and Particle Cloud into InfluxDB.
- Visualized RPM, vehicle speed, and engine load in the recovered Grafana dashboard during a recorded driving session.
- Built and tested the full path across in-vehicle hardware, cloud services, storage, and visualization.
- Preserved the original setup, console output, architecture, dashboard, and recorded-drive evidence shown here.
- C++
- OBD-II / CAN bus
- Carloop
- Particle Cloud
- InfluxDB
- Grafana
Computer vision studies
Two Python studies focused on face similarity and image-to-text extraction, including preprocessing, model integration, validation, and failure handling.
Face similarity
- Webcam
- Preprocess
- Embed
- Compare
- Rank
Image to text
- Scan
- Preprocess
- OCR
- Parse
- Review
- Face similarity study
- Webcam frames become embeddings and are ranked against a small image set using cosine similarity.
- Image-to-text study
- OpenCV preprocessing and Tesseract OCR extract contact details from scanned-card images.
- What I focused on
- Input cleanup, failed detections, model integration, validation, and readable output.
More about the work
- The need
- Personal studies focused on how computer-vision pipelines handle real inputs, failed detections, and noisy images.
- What I owned
- Input handling, preprocessing, model integration, validation, and result presentation across two study paths.
- How I built it
- Python, OpenCV, NumPy, DeepFace / FaceNet, scikit-learn, Pillow, and Tesseract OCR.
- What it produced
- The source covers two complete processing paths: ranking face similarity from webcam frames and extracting contact details from scanned-card images.
What I handled
- Worked with image files and webcam frames through Python and OpenCV.
- Generated FaceNet embeddings and compared them with cosine similarity.
- Handled missing faces, invalid inputs, and other failure paths instead of assuming a clean result.
- Used OpenCV preprocessing and Tesseract OCR in a separate image-to-text workflow.
- Documented failure cases separately from successful outputs so the behavior could be reviewed.
- Python
- OpenCV
- NumPy
- DeepFace / FaceNet
- Tesseract OCR
- Data validation
- Model testing
Skills and tools
- BuildJavaScript · HTML/CSS · Python · C++
- ConnectAPIs · Webhooks · CMS/CRM · Particle Cloud
- Data & visionOpenCV · NumPy · InfluxDB · Grafana
- ShipGit · Jekyll · GitHub Pages · QA
Web & integration
Interfaces and integrations built around the workflow they need to support.
Used inClient systems, workflow automation, and this portfolio
- JavaScript, HTML & CSSInteractive websites and browser-based experiences.
- APIs & webhooksBusiness-system and workflow integrations.
- CMS/CRM workflowsWebsites, intake paths, content operations, and customer records.
- Jekyll & GitHub PagesThe static-site and deployment foundation used for this portfolio.
- GitSource control, isolated branches, and preview workflows.
Python, data & vision
Data handling and computer-vision studies with attention to inputs and failure cases.
Used inComputer-vision studies, automation, and ML-product contribution
- PythonAutomation, data handling, and computer-vision experiments.
- OpenCV, NumPy & PillowWebcam input, image cleanup, preprocessing, and conversion.
- DeepFace / FaceNet & scikit-learnEmbedding and cosine-similarity ranking in a face-similarity study.
- Tesseract OCRText extraction from scanned-card images after OpenCV preprocessing.
- Data validation & model testingInput quality, edge cases, failure modes, and output review.
- InfluxDB & GrafanaTelemetry storage and visualization in the Arcadia capstone.
Embedded & physical systems
Code and tooling that cross the boundary between software and physical systems.
Used inVehicle telemetry capstone and Hasbro internship
- C++OBD-II requests, ECU-response processing, and Particle firmware.
- OBD-II, CAN bus & CarloopVehicle-data capture in the Arcadia capstone.
- Particle CloudMoving recorded telemetry from the vehicle setup into the data pipeline.
- Blender & 3D modelingModular modeling and production-support work during the Hasbro internship.
Implementation & delivery
The work between a technical idea and a usable release.
Used inProfessional AI implementation, client delivery, and technical support
- Requirements & workflow mappingTurn operational needs and open questions into scoped work.
- Technical project coordinationKeep dependencies, owners, status, and next actions visible.
- QA & release readinessEnd-to-end checks, issue tracking, launch material, and support planning.
- Onboarding & documentationAccess plans, handoffs, rollout guides, and feedback loops.
- Stakeholder communicationClear project updates and decision-focused handoffs.
- Zendesk & technical supportTicket triage, troubleshooting, documentation, and follow-through at Arcadia.
Education and certifications
Learning that supports the work
Education
Bachelor’s degree in Computer Science
Arcadia University
Completed 2023
Certifications
Claude Certified Architect — Foundations
Anthropic
Issued Jun 2026 · Expires Dec 2026
Building with Claude
Anthropic
Issued Apr 2026
Claude Code in Action
Anthropic
Issued Apr 2026
Contact
Get in touch.
Email me about a role or project.