Gaurav Bomra’s account of building a multi-agent orchestration system shows how a system can appear to work while quietly violating important guarantees. One test task created more than 5,000 duplicate review requests; a fallback let the service start without its required checkpointer; and replay used trace data instead of the requested checkpoint. His fixes focused on enforcing integrity in the database, failing clearly when persistence was unavailable, and keeping tracing and checkpoint identifiers distinct.
These are incidents from one engineer’s case study, not evidence of how often similar failures occur across the industry. The useful lesson for builders is narrower: test the guarantees behind a successful-looking demo, especially when retries, persistence, and replay are involved.
As an Amazon Associate I earn from qualifying purchases.
The system Bomra set out to build
In his account, Gaurav Bomra describes a supervisor agent that breaks complex work into subtasks and delegates them to specialist agents through a LangGraph state machine. The surrounding system was intended to support the full task lifecycle, not just agent responses.
PC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minute- Working state: Redis held short-term state, while ChromaDB provided long-term semantic memory.
- Human review: Sensitive actions went through a review queue and user interface for approval.
- Operations: OpenTelemetry traces covered agent calls and tool invocations; the system also tracked costs and supported replay.
- Application stack: Python and FastAPI, LangGraph, PostgreSQL, Redis, ChromaDB, Celery, OpenTelemetry, and React.
These are the architecture and capabilities Bomra reports; the account does not independently verify deployment characteristics. He introduces the story as one about “building a real multi-agent orchestration system, and three bugs that taught me more than the initial build did.” (Gaurav Bomra’s account on DEV Community)
#1 Best Overall
- Speed up your tasks with AI: Unlock new levels of productivity and creativity by upgrading to Intel Core Ultra processors with built-in AI.
- Supports multiple monitors: Connect up to four FHD monitors using DisplayPort and Daisy Chaining*. Or connect two 4K displays using HDMI 2.1 port and DisplayPort.
- Effortless upgrades: The tool-less entry and removable side panel let you quickly access the internal components, making upgrades convenient and stress-free.
- Ready for business: Keep your data secure with a hardware TPM security chip. And when you need to step away from your desk, simply secure your desktop using the built-in lock slot or padlock loop.
- Style meets sustainability: Dell Tower Desktop seamlessly combines elegance with sustainability. Its sleek, modern design, crafted from recycled materials and featuring refined corners, makes it a stylish addition to any home or office.
Three failures beyond the happy path
1. Retries created duplicate review requests
One test task generated more than 5,000 duplicate rows in the review_requests table, Bomra says. Re-evaluation during retry or timing conditions could create another pending review for the same task and subtask. The application initially checked for an existing pending request, but that check alone could not close the race: two concurrent operations could both check before either inserted a row.
Bomra added a database partial unique index, UNIQUE (task_id, subtask_id) WHERE status = 'pending', alongside the application-level check. He reports that the database constraint closed the race the application check left open. The incident illustrates why an application check and a database integrity rule serve different roles: the check can avoid unnecessary insert attempts, while the constraint is the final guard against concurrent duplicates. The reported count is a single test-task incident, not a benchmark or a measure of typical duplication.
2. A fallback let the service run without persistence
When the primary checkpointer failed to initialize, the code substituted a sentinel value and allowed the service to continue. That made startup look successful even though the required persistence layer was absent. Bomra removed the fallback so initialization failure raises a hard RuntimeError; he also added a clear success log when initialization works.
Rank #2
- 【Next-Gen AI Power & Performance 】Powered by the latest Intel Core Ultra 7-265 processor with 20 cores, 20 threads, 30 MB Intel Smart Cache, and speeds up to 5.2GHz, delivering lightning-fast responsiveness for AI workloads, creative projects, and multitasking.
- 【High-Speed DDR5 Memory & PCIe SSD Options】Choose the performance that fits your needs, from 16 GB up to 64 GB of ultra-fast DDR5 RAM and lightning-quick PCIe NVMe SSD storage ranging from 512 GB to 4 TB. Enjoy rapid file access, smooth multitasking, and plenty of room for all your projects and media.
- 【Enhanced Connectivity and Versatility】 Front port: 1 x USB Type-C (USB 10Gbps), 1 x USB Type-C (USB 5Gbps), 2 x USB Type-A (USB 10Gbps), 2 x USB Type-A (USB 5Gbps), 1 x Headphone/Microphone Combo Jack; Rear port: 4 x USB Type-A 2.0, 1 x Audio-out, 1 x Display Port, 1 x Ethernet RJ-45, 1 x HDMI; Wi-Fi 6 and Bluetooth; Wired Keyboard and Mouse
- 【HP SilentFlow Cooling】The HP SilentFlow AI hybrid cooling system automatically adjusts fan speeds and temperature levels, maintaining powerful performance with whisper-quiet operation.
- WINDOWS 11 HOME AND Microsoft Copilot - Windows 11 helps you think, express, and create in a natural way; Microsoft Copilot is always on hand to boost your productivity, accelerate your creativity, and help you communicate with maximum clarity
This changes the failure from a hidden runtime risk into a visible startup failure. It does not make the checkpointer infallible, but it prevents this particular required dependency from being silently treated as optional.
3. Replay confused a trace ID with a checkpoint ID
The frontend sent an OpenTelemetry span_id to a replay endpoint that expected a LangGraph checkpoint_id. Because lookup failed, backend fallback behavior rebuilt an approximate state from trace data rather than loading the actual checkpoint. A trace span can identify an observed operation; a checkpoint identifies saved graph state. They are not interchangeable merely because both are identifiers associated with execution.
Bomra recorded the checkpoint identifier as a span attribute and changed the frontend to send that identifier for replay. He also removed the approximate reconstruction fallback, so a missing checkpoint now produces an explicit error rather than a replay that may look plausible but is not based on the requested saved state. He says he verified the native checkpoint path using a log line; the account does not describe an independent audit of replay correctness.
Rank #3
- 14TH GEN POWER & PRO PERFORMANCE: Powered by the 14th Gen Intel Core i3-14100 processor (4-Core, 8-Thread, up to 4.7GHz Turbo, 12MB cache) and Windows 11 Pro. Built to tackle heavy business workloads, office automation, and continuous daily operations with ultra-responsive speed.
- HIGH-SPEED DDR5 & FAST NVME SSD: Equipped with a massive 512GB PCIe NVMe SSD for storing large database files, media archives, and projects with ease. Combined with 8GB high-speed DDR5 RAM to eliminate lag during heavy, multi-application processing.
- 4K MULTI-MONITOR SUPPORT: Intel UHD Graphics 730 supports up to dual 4K monitors via HDMI 2.1 and DisplayPort 1.4a. Ideal for financial trading, content previewing, and complex data analysis requiring vast visual real estate and crisp clarity.
- COMPREHENSIVE CONNECTIVITY & PORTS: Next-gen MediaTek Wi-Fi 6 and Bluetooth ensure seamless wireless performance. Fully equipped with modern ports including USB 3.2 Gen 1 Type-C, USB-A, HDMI 2.1, DisplayPort 1.4, RJ45 Gigabit Ethernet, SD media reader, and audio jack.
- ENTERPRISE-READY & OPTIMIZED DESIGN: Pre-loaded with Windows 11 Pro 64-bit for enterprise-grade security and IT manageability. Features a sleek, space-saving desktop footprint (12.76" x 6.06" x 11.53") designed with an optimized thermal airflow layout for system longevity.
What these bugs suggest for agent-system design
Put concurrency guarantees where concurrency is enforced
If duplicate pending work would be harmful, make uniqueness a database invariant as well as an application behavior. An application-level “check, then insert” sequence can race under concurrent requests. A database constraint can reject the conflicting write even when two callers pass their checks at nearly the same time. The right constraint depends on the actual business rule; Bomra’s example applies uniqueness only to rows whose status is pending.
Recommended Free Tools
Make required dependencies fail visibly
A fallback is useful when degraded operation is genuinely safe and clearly communicated. It is dangerous when it converts a required capability into an invisible omission. In this case, allowing startup without the checkpointer concealed the absence of persistence. For required state infrastructure, failing initialization explicitly can be safer than letting later operations proceed under a false assumption.
Keep identifiers and data paths explicit
Observability data can help explain execution, but a trace is not automatically a substitute for the state a replay endpoint promises to load. Define which identifier each API accepts, pass the correct one through the UI, and make missing state an explicit error. A fallback that approximates data should be clearly separated from a true checkpoint replay rather than presented as equivalent.
Rank #4
- Built for Local AI and Advanced Workflows – The BOSGAME M5 AI Mini PC is powered by AMD Ryzen AI Max+ 395 with 16 cores, 32 threads, up to 5.1GHz, 50 TOPS NPU performance and up to 126 TOPS total AI performance. It is designed for local AI inference, private AI assistants, coding, data analysis, virtualization, content creation and demanding multitasking while keeping sensitive data on the device.
- 128GB Unified Memory for Large Models and Creative Projects – M5 includes 128GB LPDDR5X-8000 unified memory, giving the CPU and Radeon 8060S graphics access to a large shared memory pool. This helps support memory-intensive AI workloads, large project files, multiple virtual machines, 3D work, video editing and complex professional applications without the capacity limits of typical 32GB or 64GB mini computers.
- Radeon 8060S Graphics for Creation, Rendering and Gaming – Integrated Radeon 8060S graphics with 40 RDNA 3.5 compute units delivers high-end visual performance without a separate graphics card. Use the M5 creator workstation for 4K video editing, 3D rendering, CAD, AI image workflows, high-resolution media and modern gaming, while maintaining a compact desktop footprint.
- 2TB PCIe 4.0 SSD and Flexible Expansion – A pre-installed 2TB NVMe PCIe 4.0 SSD provides fast access to models, datasets, media libraries and project files. A second M.2 2280 PCIe 4.0 slot allows additional storage expansion, while the SD 4.0 card reader supports efficient photo and video workflows for creators and production teams.
- Professional Connectivity and Four-Display Support – Dual USB4 ports, HDMI 2.1 and DisplayPort 1.4 support up to four displays and resolutions up to 8K@60Hz. WiFi 7, Bluetooth 5.4 and 2.5GbE deliver fast networking for cloud collaboration, NAS access and business deployment. Windows 11 Pro, performance-mode switching, Wake-on-LAN and auto power-on support flexible workstation use.
Test guarantees, not just visible success
The three incidents share a pattern in Bomra’s telling: the system could appear to proceed while an important guarantee had failed. A test or demo that confirms an answer appears, a service starts, or a replay screen loads may not prove that writes are unique, persistence is active, or the requested checkpoint was restored. Tests should exercise those guarantees directly, including concurrent creation, initialization failure, and replay with a missing or mismatched checkpoint.
Why the account is useful—and what it does not establish
Bomra’s practical warning is: “A fallback that masks a real problem is worse than no fallback at all.” That is his lesson from these incidents, not a universal rule that every fallback should be removed. A fallback can be appropriate when degraded behavior is intentional, bounded, and visible; the cases here concern fallbacks that obscured the loss of a required property.
Quick wins for a faster PC:
Scan for outdated or missing drivers - takes under a minuteDriver Scan →Repair Windows errors before they cause bigger problemsFix Now →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →The DEV Community page shows a Sep 21 posting date but does not display a year. It provides one engineer’s account and no independent incident report, repository audit, population-level failure rate, or comparative benchmark. Its strongest contribution is therefore concrete: it shows how duplicate side effects, missing persistence, and misleading replay can arise in a single orchestration system, and describes the changes its author says addressed them.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
