Planned outline · Module 01
Web research infrastructure
Search indexes, crawling, extraction, RAG, provider selection and citations.
Practice: Draw a research request from browser to verified evidence.
Builder Academy · Learning pathway
Explore how software, data, research, security and operations fit together. Start with the mental model, try a synthetic exercise, then apply it to a portfolio project.
Planned outline · Module 01
Search indexes, crawling, extraction, RAG, provider selection and citations.
Practice: Draw a research request from browser to verified evidence.
Planned outline · Module 02
Modular monoliths, service boundaries, coupling and scaling.
Practice: Choose a boundary for a new research consumer.
Planned outline · Module 03
Sources, ETL/ELT, transformations, contracts, versioning and quality.
Practice: Trace one RSS item through a pipeline.
Planned outline · Module 04
HTTP, REST, GraphQL, webhooks, authorization, pagination and versioning.
Practice: Design an authenticated research API contract.
Planned outline · Module 05
Parallelism, asynchronous jobs, queues, retries, consistency and idempotency.
Practice: Find duplicate work in a retried job.
Planned outline · Module 06
Threat modeling, least privilege, SSRF, injection, secrets and supply-chain safety.
Practice: Identify two unsafe trust boundaries.
Planned outline · Module 07
Taxonomies, ontologies, entity resolution, relationships and temporal history.
Practice: Resolve two names to one vendor entity.
Planned outline · Module 08
Structured logs, metrics, traces, SLOs, alerts and incident response.
Practice: Diagnose a failed extraction task.
Planned outline · Module 09
RAG, embeddings, reranking, tool calls, MCP and evaluation.
Practice: Verify a machine-generated claim.
Planned outline · Module 10
Environments, Docker, CI/CD, deployments and rollback.
Practice: Plan a safe database migration.
Planned outline · Module 11
Provider prices, latency, cache, usage budgets and build-versus-buy.
Practice: Compare cost per verified finding.
Planned outline · Module 12
ADRs, ownership, reviews, versioning and deprecation.
Practice: Write an architecture decision record.
Planned outline · Module 13
Discovery, PRDs, roadmaps, UX, accessibility and feature flags.
Practice: Define acceptance criteria for a lab.
Specifications only: these labs are not yet interactive or connected to paid APIs.
Frontend → API → queue → provider → database; identify failure boundaries.
Collect → validate → deduplicate → normalize → store → review.
Compare sequential work, concurrency, retries and idempotency.
Identify prompt injection, SSRF and access-control weaknesses using synthetic examples.
Investigate a fictional provider outage, leaked token or deploy failure.
Compare mock Exa, Firecrawl and Parallel responses; never label fixtures as live measurements.
Inspect citations, source hashes, freshness, conflicts and review status.
Use public or synthetic cybersecurity sources to map source → capture → hash → entity → claim → review → downstream product. Never confuse a provider result with independently verified evidence.
Start with the Account Signal Engine Research Hub and its RSS / Firecrawl foundation; add paid providers only after security, cost and quality evaluation.