Senior engineering team forB2B SaaS, platforms, web,mobile and DevOps.
BFG.AM helps product companies and digital agencies ship product modules, backend services, infrastructure work and mobile apps — without long hiring cycles or freelancer chaos.
BFG.AM is the engineering brand of BFG LLC, Armenia.
When your roadmap grows faster than your hiring pipeline
Hiring senior engineers takes months. Freelancers are hard to manage. Internal teams are already busy with core product work. Product modules, backend services and infrastructure tasks keep piling up.
BFG.AM provides a managed senior team that takes ownership of a bounded scope, integrates into your workflow and delivers with clear communication.
What we do
Three engagement formats covering product, platform and white-label delivery.
Product Engineering Team
PM-managed team for SaaS modules, dashboards, admin panels, customer portals, MVPs and mobile applications.
Platform & DevOps Team
Go and backend services, APIs, integrations, CI/CD, Kubernetes, cloud infrastructure and observability.
White-label Delivery for Agencies
Senior engineering capacity behind your agency brand — web, mobile, backend and DevOps delivery without expanding your in-house team.
Built for product teams under pressure
A few common situations we step into.
CTOs & Heads of Engineering
Cover roadmap gaps without waiting months to hire.
Product teams & founders
Take product modules from backlog to production with a managed team.
Digital agencies
Add reliable engineering delivery behind your brand.
2-week paid pilot. Money-back if we miss expectations.
The lowest-risk way to see how we ship before any long-term commitment. At the week-1 demo you decide: continue or full refund — no questions.
- →A bounded module — what a senior engineer would ship in two weeks
- →1–2 senior engineers and a PM running the sprint end-to-end
- →Daily progress updates, midpoint demo, final demo and handoff
- ✓Working code merged into your repository
- ✓Technical audit and risk map of the surrounding system
- ✓30 / 60 / 90-day delivery roadmap recommendation
A real team, not a freelancer pool
We have already worked together on long-term production systems. Clients get engineers, delivery management and technical ownership — not a random set of contractors.
Engineering teams already shipping with us.
BFG embedded a senior frontend team and a designer into our virtualization platform. Six months in they were shipping consoles in production at our own team's velocity.
We needed mobile, IoT backend and an operator CRM for a parking-sharing platform — without spinning up an in-house team. BFG delivered all three under one PM.
BFG runs the engineering infrastructure behind our enterprise mail platform — seven bare-metal hosts, full HA, in production since August 2025.
Start small, keep control, scale when it makes sense.
A predictable rhythm from first call to delivery.
- 01
Discovery call
We listen to the problem, the scope and the constraints. 30–45 minutes.
- 02
Scope and team setup
We propose the smallest useful team and a bounded first sprint.
- 03
Sprint delivery
PM-managed delivery with clear ownership and a working baseline.
- 04
Weekly demo
Live demos, written progress notes, no surprises.
- 05
Handoff or scale
Either we hand off cleanly or extend the team for the next phase.
Selected work
Production systems we have shipped — B2B SaaS, enterprise tooling and infrastructure platforms.
Datacenter Virtualization Console
Web management console for an enterprise virtualization platform — the vCenter / vSphere counterpart to our hypervisor console work for the same customer. Multi-cluster topology, host inventory, VM lifecycle, storage and network policy across multiple sites. Design owned by our design team lead Anri Marutyan and built by the full BFG frontend team in Vue 3 / Nuxt 3 / TypeScript, integrated to the customer's platform backend over REST and WebSockets. Keeps the operational mental model of vSphere administrators intact, with a distinct visual system and modern interaction patterns. In production since late 2024.
Hypervisor Management Console
Web management console for an ESXi-class single-node hypervisor — the per-host counterpart to our datacenter virtualization console for the same customer. Design owned by our design team lead Anri Marutyan and built by the full BFG frontend team in Vue 3 / Nuxt 3 / TypeScript, integrated to the customer's hypervisor backend over REST and WebSockets. The interface keeps everything where existing virtualization operations teams expect it (so they do not have to relearn), but with a distinct visual identity and a more modern interaction model. In production since late 2024.
SPICE Multi-Platform Client Suite
A full suite of SPICE clients delivered for one enterprise virtualization vendor — built from scratch, not based on the open-source HTML5 reference. Four parallel surfaces ship together: a browser-based client with no plugin install, a single Go codebase native client across Linux, macOS and Windows on amd64 and arm, native iOS and Android clients optimized for touch and cellular networks, and Windows / Linux guest utilities bringing vGPU and 3D acceleration (DirectX, OpenGL, Vulkan) into SPICE for CAD and engineering VDI workloads. End-to-end UX/UI delivery and onboarding across all four surfaces.
Metro Driver Mobile App
Flutter mobile application used daily by 2k+ drivers of a metropolitan transit authority — single codebase across iOS and Android, built by mobile team lead Karen Simonyan with design by team lead Anri Marutyan. Drivers see their published shift schedule, declare overtime, register attendance and receive push notifications about schedule changes — including signal-loss aware behaviour in metro tunnels: last-known schedule is cached and write-behind synced on reconnect, so a driver loses no context underground. Talks to the platform backend over REST for synchronous reads and gRPC for streamed updates, and receives Kafka-driven push events via FCM and APNs. In production from 2023 through 2025.
Metro Driver Dispatcher Console
Operations console for the depot dispatchers who manage the shift schedules and overtime of 2k+ metro drivers. Design by team lead Anri Marutyan, frontend built by the BFG frontend team led by Roland Simonyan in Vue 3 / Nuxt 3 / TypeScript. Dispatchers compose, edit and approve driver schedules across multiple depots simultaneously, with optimistic versioning that resolves conflicting edits when two dispatchers touch the same schedule. Real-time updates over WebSockets keep all open sessions in sync. Integrates with the same Go backend platform that powers the mobile app and feeds the platform-level hand-off to HR and accounting systems. In production from 2023 through 2025.
Metro Driver Backend Platform
Distributed backend platform for a metropolitan transit authority — 10 Go services organised as a multi-module workspace (go.work), in production from 2023 through 2025. Service-to-service messaging is gRPC over Protocol Buffers (shared proto with buf generate + lint); external HTTP endpoints for the web dispatcher console and the mobile app are described by OpenAPI and generated with ogen. Storage is PostgreSQL 14 with pgx/v5 and goose-managed migrations. Async fan-out is Apache Kafka in KRaft mode (no ZooKeeper), Sarama client, at-least-once with idempotency enforced by entity-state checks. Clean Architecture is the per-service skeleton — DTO / Input / Model / Record layers with explicit converters, DIP-style interfaces declared by consumers, lazy DI in internal/app/di.go. DDD for the harder corners (overtime calculation, shift state machine): rich entities, Value Objects, and a clear domain-vs-application service split. Built by Simon Lapin (Principal Architect) with senior Go engineers Denis Stashkov, Ilya Karasev, Anna Yakubovskaya and Alexey Klimko.
Building Heating Automation & Dispatching
A building automation system controlling 372 infrared heating panels across a 5000 m² site. Replaced a passive SCADA integration with an active control service: ingests data from temperature sensors, relay cabinets and outdoor weather, factors in historical heating-and-cooling inertia per zone, and drives per-room heating decisions. Later integrated with the building's motion sensors (originally part of the security system) to add predictive control — do not heat zones with no people at night, prioritize heating where occupants actually are or will be. Includes a real-time dispatching dashboard for facility operations: per-panel status, per-zone mode, live energy consumption. Beckhoff PLC integration; in production since January 2026.
Cross-Network Social Graph Service
A Go microservice that bridges multiple identity and social-graph providers (Facebook, Google and several regional networks) into a single in-platform friend graph for a specialized designer-focused social network (Pinterest-class). Handles OAuth integration with each provider, normalizes their friend-graph schemas and runs cross-provider matching: 1st-degree friends (people the user already knows on at least one external network who also use the platform) and 2nd-degree friends-of-friends used as the basis of the in-platform recommendations engine. In production since mid-2025.
Virtualization Inventory Engine
A custom Go inventory engine built to fix the tree-rendering and reporting bottleneck in a vCenter-class virtualization platform. Stores typed objects — clusters, datacenters, hosts, VMs, datastores, networks — in a structured graph-like index with compression, caching and deduplication built into the storage layer. Used as the backbone for every operation that touches platform topology: tree rendering, search, ad-hoc reports ("how many VMs in this datacenter?", "what datastores does this VM use?"), permission checks. Scales to hundreds of thousands of objects without the linear-time scans that previously slowed the backend.
Enterprise Mail Mobile Client (iOS / Android)
Native iOS and Android mobile mail client for the same Exchange-replacement enterprise mail server we built the admin console for. Designed for end-users at scale: inbox, threading, push notifications, calendar invites, attachments and the offline behaviour mobile mail users actually rely on. First versions in production, with new functionality added iteration by iteration. Built by mobile team lead Karen Simonyan with design input from Anri Marutyan — single Flutter codebase across iOS and Android, native shims where the platform requires them.
Parking Sharing Mobile App
Consumer-facing mobile app for an hourly parking-sharing platform serving private residential complexes. Designed by our design team lead Anri Marutyan and built by mobile team lead Karen Simonyan: search nearby available spots, book by the hour, drive in (the electronic gate blocker unlocks via the booking-aware backend), pay through the integrated checkout. Single Flutter codebase across iOS and Android. One of three sibling cases delivered as a turnkey project team for this customer.
Parking Sharing — IoT Backend & Payments
Go backend for an hourly parking-sharing platform serving private residential complexes. Owns the booking lifecycle, real-time control of electronic parking blockers over MQTT, usage tracking with over-stay and re-use detection, and the billing/settlement flow. Integrated with a payment provider for in-app acquiring and provider-side payouts to individual spot owners. Designed for residential complexes where parking spots are individually owned but shared by the hour. One of three sibling cases delivered as a turnkey project team for this customer.
Parking Sharing — Operator Console & CRM
Operator-side admin console and customer-relationship system for the parking-sharing platform. Lets the operator manage spots and blockers, run statistics and revenue dashboards, track customer history and payment status, and handle support cases. Design by our design team lead Anri Marutyan, frontend by the BFG frontend team in Vue 3 / Nuxt 3 / TypeScript, integrated to the same backend that powers the mobile app. One of three sibling cases delivered as a turnkey project team for this customer.
Enterprise Mail Server Admin Console
Admin console and operations dashboards for an enterprise mail server in the Exchange-replacement class. Design from scratch by our design team lead Anri Marutyan: information architecture for mailbox, distribution group, transport rule, retention policy and queue management surfaces, plus the operations dashboards (queue depth, message flow, server health, storage). Full frontend implementation by the BFG frontend team in Vue 3 / Nuxt 3 / TypeScript, integrated over REST and WebSockets to the customer's mail backend. Heavy on table-and-form discipline — every admin surface needs to handle thousands of entities without freezing or losing context.
End-to-End Engineering Org Infrastructure
Full company infrastructure delivery for an engineering organization — from servers and storage to every service the team uses day-to-day. Designed and deployed on 7 HP bare-metal servers running Proxmox with a Ceph distributed storage cluster, fronted by a Keycloak SSO that owns the central user directory and authenticates every internal service — including the WireGuard VPN. Engineering stack: GitLab + GitLab Runners + Docker Registry, YouTrack issue tracking, MinIO object storage, Mattermost for team chat, knowledge base, forum and documentation platform. Back-office: CRM and ERP systems on the same infrastructure. Backups with dedicated storage, HA across the cluster, VPN-only external access. In production since August 2025, currently serving 240 daily users.
Hospital Pharmacy System — Legacy Modernization
A hospital pharmacy management system tracking medication dispensing, dosage variants (half pills, 1.5 pills, custom doses) and controlled-substance prescriptions. Written decades ago, hard to extend, operationally critical and not replaceable in a single pass. We containerized the legacy application with no source rewrites, packaged it with sidecars for resource reservations and operational concerns, and deployed it on Kubernetes. Set up modern CI/CD, documented the infrastructure for the client's engineering team, and defined integration points so new modules can grow alongside the legacy core in modern languages — a strangler-fig path to gradual replacement without taking the system offline.
VM Cluster Filesystem
A cluster pseudo-filesystem for virtualization workloads, designed as a VMFS-class equivalent for shared block storage. Uses a vector-clock-based DLM written through the underlying disk on storage that does not expose SCSI reservations; falls back to SCSI reservations or ATS (hardware-assisted locking) primitives where the array supports them; uses Ceph's native locking when running on Ceph. Backends: LVM block storage or Ceph. Supports thin volumes, snapshots and live VM migration. Production-validated at 32 nodes per cluster, 14000 VMs and 500 concurrent disk expansions.
Auto Parts Analytics & Cross-Reference Engine
A Go backend powering an automotive aftermarket marketplace. Ingests 3600+ parts catalogs from manufacturers and distributors, normalizes them and builds a cross-reference graph across part numbers and alternative suppliers. Buyers search by any part number and get every compatible alternative ranked by price, quality and purchase popularity. Operators see seasonal demand and stock ahead of price curves. Sustained 1900 requests per minute at peak. Heavy engineering on storage layout, indexing strategy and API latency.
Our tech stack
Battle-tested technologies we trust in production.
Pick the smallest format that solves the problem.
Start with one specialist, a bounded sprint or a small managed team. We keep the setup practical and easy to control.
- Dedicated engineer
- Managed product team
- Platform / DevOps team
- Fixed-scope sprint
- White-label delivery partner
Send us a short brief or backlog.
We will suggest the smallest useful team or sprint to start with.