Lynon Solutions

Available to hire

Lynon Solutions is a software agency. We design iOS, Android and web products.

About

Lynon Solutions is a software agency. We design iOS, Android and web products. We build websites, web apps, SaaS and AI agents. For example, we built a multi-tenant monitoring SaaS with tested tenant isolation.

WHAT WE DO

  • UI/UX and product design for iOS, Android and the web.
  • Websites and landing pages, including sites in two languages.
  • Web apps and SaaS. The database controls who can see which data.
  • AI agents, RAG assistants that cite their sources and voice agents for phone calls.

HOW WE WORK
We use an AI-assisted workflow under written rules. We own the architecture. Tests check access rules and architecture rules before we release a change. You get a short plan first. We report progress and risk in plain words.

GOOD FIT

  • You want one team to design and build your product.
  • You need design first and development later.
  • You have an AI prototype that works in a demo but fails in real use.
  • Joined September 2026

Languages

  • EnglishFluent

Industry experience

  • Software & Internet
  • Education
  • Travel & Hospitality
  • Retail

More experts in Istanbul, Turkey

Projects · 29

Multi-Tenant AI Support SaaS with Database-Enforced Security

This is an AI support platform for businesses. One agent answers across six channels from one workspace. There is no long-running app server. All server logic runs in edge functions, which run per request and keep no state. The database enforces access. The browser queries PostgreSQL directly. Its row-level security policies decide which rows each business can see. We turned ten architecture rules into automated tests. Code that breaks a rule stops the build. Another test compares schema changes in the live database with those in the code. A gap in either direction stops the build.

  • Full Stack Developer
  • Web Developer
Model-Agnostic Voice and Chat AI: Provider Routing with Fallback

This is a conversational AI platform for a confidential client. We set up its AI engineering from the start. We wrote its agent runtime, the code that runs its AI agents. The platform does not depend on one AI provider. One component selects the language model, speech-to-text and text-to-speech providers. All 3 have ordered backup lists. We chose not to use an off-the-shelf LLM gateway, a service between the app and the model providers. Its main feature is automatic retry on another model. We kept the retry choice in the agent runtime, which sees facts that a gateway cannot see.

  • Full Stack Developer
  • Web Developer
Replacing a Live AI Pipeline Without Downtime

This is an AI assistant product on web and mobile, across 5 repositories. We owned the architecture. We wrote the specifications that other engineers built from. We added a new system that runs AI agents beside the live AI pipeline, instead of a rewrite. Switches choose the old pipeline or the new path. The new path has its own database. We removed the canary and the percentage rollout. Both send a change to some users first. We replaced them with one switch. The new path stays off unless a second safety switch, the grounding gate, is also on.

  • Full Stack Developer
  • Web Developer
Real-Time Phone Voice Agent: Fixing Silence and Latency

We built a voice agent that answers real phone calls. Each call opened with silence while the speech services and language model started. The greeting never changes, so we generate its audio once, when the program starts. We store that audio in memory as 20 ms chunks, ready to send when the call connects. A second bug silenced every call after the first. Some speech services store session state: data tied to one call. Reusing them breaks the next call. Only the language model is safe to reuse between calls.

  • Full Stack Developer
  • Web Developer
Multi-Tenant Monitoring SaaS with Tested Tenant Isolation

This is a multi-tenant SaaS that watches TLS certificates and warns before they expire. It has tenant isolation, billing, CI (automatic checks on each change) and end-to-end tests. We built it with our AI-assisted workflow, under written rules. We own the architecture. Row-level security policies in PostgreSQL enforce tenant isolation inside the database. Row-level security tests and end-to-end tests run on each change.

  • Full Stack Developer
  • Web Developer
RAG Tutor: Changing the Embedding Model on a Live System

This retrieval-augmented tutor finds source passages before it answers. Each answer cites its source passage. An embedding model turns text into a vector: a list of numbers. Each number is one dimension. Our vector index stores vectors for search, up to a dimension limit. The larger model we wanted went over that limit. We used it at reduced dimensions instead of a smaller model. The live migration used a shadow column for the new vectors and a second search function. We wrote the rollback script before the migration ran.

  • Full Stack Developer
  • Web Developer
LLM Evaluation You Can Trust: Catching a Harness That Lied

Our evaluation of the retriever, the AI's search step, reported low recall. Recall is the share of correct records that the retriever returns. But the retriever returned every correct record. The test query did not apply the production visibility rule. So the evaluation expected records that the retriever could not see. The run showed no error. A low number like this sends a team to fix a system that works. A separate evaluation checks model and prompt changes against a fixed test set in version control. Each run must pass written score limits. Code grades each output by exact match. No AI model grades the output.

  • Full Stack Developer
  • Web Developer
Autonomous Publishing Agent with Deterministic Safety Gates

This agent researches, writes and publishes on a schedule, with no human review. The design goal is to make it safe to run unattended. Every claim must come from a source that we can cite. The agent builds each post from those claims and adds none of its own. A deterministic safety gate checks each post before the agent publishes it. It runs fixed code rules, so the same post always gets the same result. It blocks banned names, known false figures, placeholders and business numbers. No AI model decides what the gate blocks.

  • Full Stack Developer
  • Web Developer
Bilingual Website Prototypes from One Typed Config

Small businesses in real estate, law, retail, hospitality and consulting needed websites in two languages. Each website started from a short brief. Building each website repeated the same work. We moved that work into one TypeScript configuration file. TypeScript defines the shape of its data. Each website gets its own theme through design tokens: named style values such as colors. Each piece of content text is a pair in both languages. In React, every page section reads its text from these pairs. Most new content is a data change.

  • Full Stack Developer
  • Web Developer
Halyard: Sales CRM Web App with a Pipeline Board

Product design by Lynon Solutions. Halyard is a sales CRM web app with a pipeline board. It is for B2B sales teams and their managers. The home page shows quota progress, open deals by stage and the tasks for today. One progress line shows the team quota on the home page and the stage history on each deal. Each deal card shows its days in the current stage, so idle deals are easy to find.

  • Product Designer
  • UI Designer
  • UX Designer
Lentisk: Direct Booking Website for Small Hotels

Product design by Lynon Solutions. Lentisk is a booking website for small hotels that want guests to book directly. Guests compare four kinds of room, pick dates on a calendar and book on the hotel's own site. We put the booking bar on the home page, right under the hero photo. On the room page, the calendar shows the nightly rate under each date.

  • Product Designer
  • UI Designer
  • UX Designer
Kiln: Strength Training Planner for iOS

Product design by Lynon Solutions. Kiln is an iOS app that plans strength workouts and records each set. It is for lifters who follow a program and track their weights. The app shows which muscles need more rest and which are ready to train. We used one color scale for recovery. Red marks a muscle that the user trained recently. Pale green marks a muscle that is ready. One screen holds the set log, the rest timer and the plates to load on the bar.

  • Product Designer
  • UI Designer
  • UX Designer

Work experience

Design and Development Team · Lynon SolutionsSep 2017 – Present · 9 yrs 3 mos

We design iOS, Android and web products. We build websites, web apps, SaaS and AI agents. One team owns the work from design to release.

Hire Lynon Solutions today

Post a job and invite Lynon, or contact them directly. No hidden commission.

Post a job