Senior software engineering · Vancouver, BC · Making software since 1994[email protected] · US and Canadian business hours

Senior engineering firm · Vancouver

You built it with AI. We help you make it hold.

Founders now build real products with Lovable, Bolt, Replit, Cursor, Base44 and v0, and get further than ever before they need an engineer. DIGI117 is the senior engineering firm for the point where they do: a stuck app, a first version that has to hold up for real users, AI features that cost more than they earn, or a technical partner for the company. Every engagement is led by senior engineers and reviewed by our founder, Vadym Chernega, who has been making software since 1994. AI agents do the typing. An engineer reviews every change, and DIGI117 answers for the result.

Get a free first look Looking for a technical co-founder?

BUILT WITH AI Sign-in Payments Deploys Data rules HOLDS UP Tested sign-in Live payments Clean deploys Checked access FIRST LOOK THE FIX YOU KEEP IT
1994Making software since. BASIC on the ZX Spectrum to today's AI agents.
FreeFirst look at a stuck app, in writing, from a senior engineer.
1 dayTo a written answer from a person. You talk to the people doing the work.
30 daysWarranty on every fix. If it breaks as built, we fix it at no charge.
Tools and platforms we fix and build on
  • Lovable
  • Bolt
  • Replit
  • Cursor
  • Base44
  • v0
  • Supabase
  • Firebase
  • Stripe
  • React
  • Next.js
  • Node.js
  • Python
  • PostgreSQL
  • AWS
  • Cloudflare

What we do

From the first wall to the first engineers

Each service starts with a short first step, so you see how we work and what the full job costs before you commit to it. The first look at a stuck app costs nothing.

01

Unstuck

You built an app with Lovable, Bolt, Replit or Cursor, and the next step will not work: sign-in, payments, a deploy, data that leaks or disappears, or a loop that burns credits. Send the link. A senior engineer replies within one business day with what is wrong and what it takes, and if you can fix it yourself, how.

First look freeEach fix quoted before work starts
Unstuck
02

Co-founder or CTO

For founders who own the market and need someone to own the technology: a technical co-founder, a part-time CTO or an advisory board seat. We take a small number, by application, with the terms in writing.

Cash first, equity on topBy application
Co-founder or CTO
03

MVPs

A first version real users can sign up for, in 8 to 12 weeks, built new or finished from a prototype you started. The scope is fixed in writing after a paid scoping sprint, and you see a working demo every week.

Three-day scoping sprint $4,500Most MVPs $45,000 to $90,000
MVPs
04

AI in your product

Assistants, agents, search and document features users will pay for. We measure the cost per user and the quality on a test set before launch, so the model bill does not grow faster than the users.

AI map $5,900Pilots from $15,000
AI in your product
05

Custom software

Internal tools, client portals and integrations that join the systems you already run. Built to be operated and changed by your people after we leave, with tests, documentation and a runbook.

Custom builds from $25,000
Custom software
06

Senior engineers

Named senior people inside your team, for as long as you need them. They join your stand-ups, work in your repositories and follow your process. Every invoice shows who worked, at what level, on what.

From $150 an hour
Senior engineers

Why companies bring us the hard ones

Senior by training. Fast by method. Accountable in writing.

Plenty of firms say senior. Ours learned the trade below the framework, and still read the code before they quote.

Trained from the metal up

Our engineers learned on infrastructure, low-level code, backends and data before they learned frameworks. Generated code holds no mystery for people who have read every kind of code since it was new.

  • Assembler to TypeScript, CGI to serverless
  • Stored procedures, nightly batch, message queues
  • Postgres, cloud infrastructure as code

Agents type, engineers decide

Parallel AI agents write code and drafts from a written spec. A senior engineer reviews every diff, and no agent merges its own work.

  • Written spec and acceptance list per task
  • Tests before trust, an eval gate for AI features
  • A person on every merge button

Everything in writing

Prices on the website. Quotes before work starts. Decisions written down where you can read them. If you can fix it yourself, we say so on the first look, even when it means no job for us.

  • NDA before scoping
  • Weekly demo, shared channel, written status
  • Itemised invoices by person, level and task

How an engagement runs

Read the code first. Then a price. Then weekly proof.

The same four stages for a fix, an MVP or a build. You can stop after any of them and keep everything we produced.

Days 1 to 5

Read and inventory

We read the code, the data and the infrastructure, or work the idea through with you. You get a written inventory and a list of risks, ranked.

Week 2

Fixed quote

Scope, price, schedule and the people on the job, in writing. Changes to scope get their own written quote before anyone builds them.

Every week

Build and demo

Working software every week, in your repository. Ten-minute stand-ups, a shared channel, and a written status you can forward to your board.

Handoff

Hand over, cleanly

Code, tests, documentation and a runbook in your accounts. Your team can run and change it without us. Support continues only if you want it.

Selected work

How we decide, shown in full

A complete case study with the research, the numbers and every cut, plus systems we built that are in use today. Clients are not named unless they agree to it.

Case study · Research to MVP plan

Peoplarium: a verified meetup platform, from research to price

How we researched, cut, designed and priced a meetup platform for Vancouver. Each finding has its source and date, the business model is a calculator you can change, and the architecture is mapped part by part.

28research tasks in 4 waves
3plan versions, each cutting more
~13weeks to launch
4,600members to break even
Read the case study

CRM · AI agents

Agentic CRM for a services sales team

Agents research, qualify and prepare. People decide who gets a call. Runs on the client's own servers.

In daily use

AI assistant · Leadership

Assistant for an executive's inbox and calendar

Triage, reply drafts, meeting briefs and follow-ups. The assistant prepares; a person sends.

In daily use

MVP · Hardware market access

Market-access lookup for imported hardware

Buyers check a robot or device by brand, model or serial number: where it can be sold and whether the unit is genuine.

Live in five languages

All work

Engagement models and pricing

Prices on the page, quotes in writing

Four ways to work with us. Every job is quoted in writing before work starts. All prices in US dollars.

ModelBest forHow it is pricedFirst step
UnstuckAn app built with an AI tool that hit a wallFixed quote per fix, before any work starts. Most fixes cost a few hundred to a few thousand dollars.A free first look, in writing, within one business day
Fixed-price projectMVPs, custom software, AI pilots with a clear goalFixed price per agreed scope. Changes quoted separately, in writing.MVPs: a $4,500 scoping sprint. AI: a $5,900 map.
Co-founder or CTOFounders who own the market and need a technical partnerCash first, with a small vesting stake on top. Terms per company, in writing.An application, then a call
Senior engineersTeams that need senior capacity inside their own processFrom $150 an hour, billed monthly, itemised by person, level and task.A short brief and an intro call with the engineer

Large programs that need more people run together with our sister company, Intersog, under one written agreement.

Since 1994

We have read code since before it wrote itself.

Every language, database and platform from the ZX Spectrum to today's AI agents. Whatever your AI tool generated, it learned from code like ours.

1994

First lines

BASIC on the ZX Spectrum, Assembler on early x86 PCs and their 640K.

1990s

Desktop and early web

C, C++, Pascal, Perl CGI, HTML from before CSS existed.

2000s

Enterprise

Java, .NET, PHP, Oracle, SQL Server, stored procedures, nightly batch.

2010s

Mobile and cloud

Swift, Kotlin, Python, Node.js, PostgreSQL, MongoDB, AWS.

Now

Agents

TypeScript, React, Kubernetes, Cloudflare, vector search, AI agents with a person reviewing every diff.

Same habit, thirty-two years apart: nothing ships unread.

For your procurement team

Paperwork your legal team will recognise

DIGI117 Ltd. is a British Columbia company. We sign before we scope, and we write down how we handle code, data and AI.

Read the Trust page

  • NDA before scopingYours or ours, signed before we see code or data.
  • Yours from the first commitCode lives in your repositories from day one. IP transfers to you in full on final payment.
  • Data stays in your accountsWe work in your cloud and repositories. Model calls go only to vendors named in the proposal.
  • No training on your dataWe use vendor terms that exclude training on client data.
  • 30-day warranty on every fixIf it breaks as built, we fix it at no charge.
  • Same-day noticeIf we find a security issue in your system, you hear it from us that day, in writing.
  • One company, one contractDIGI117 answers for everyone it puts on your project. Large programs run with Intersog under one written agreement.
  • Documents on requestSample MSA, security questionnaire answers, a DPA, and insurance certificates through Intersog.

Before the first call

How fast can you start?

The first step starts once the NDA is signed and we have read access to the code. If we cannot start when you need us, we say so on the first call, not after the contract.

Who will actually work on our project?

Named senior engineers, introduced before the quote is signed. The founder reviews every engagement. When a program needs more people, it runs together with our sister company, Intersog, under one written agreement.

How do you use AI, and is our code safe with it?

Agents write code and drafts from a written spec, inside your accounts. A senior engineer reviews every change before it merges. We use model vendors named in the proposal, on terms that exclude training on your data.

What if the job turns out bigger than quoted?

We stop and tell you in writing before doing unquoted work. You get the reason, the options and a price for each. Nothing is billed that you did not approve.

Start a conversation

Tell us what is stuck.

A senior engineer reads every message and replies within one business day with a first view of the problem and the right first step. No sales sequence.

Get a free first look Something else