https://bydtech.in/blog What changed in AI, apps and the web, releases of our open-source tools, and engineering deep dives. en-us Fri, 06 Nov 2026 00:00:00 GMT https://bydtech.in/blog/tokenization-explainedhttps://bydtech.in/blog/tokenization-explainedFri, 06 Nov 2026 00:00:00 GMTDeep divesHow LLM tokenizers work: BPE, byte-level BPE, WordPiece and Unigram, vocabulary size, and why one text costs different token counts across models. https://bydtech.in/blog/voice-agent-architecturehttps://bydtech.in/blog/voice-agent-architectureFri, 06 Nov 2026 00:00:00 GMTDeep divesHow voice agents work: cascaded STT, LLM and TTS versus speech-to-speech, the latency budget per turn, turn detection, barge-in, telephony and evaluation. https://bydtech.in/blog/mcp-server-transports-and-authhttps://bydtech.in/blog/mcp-server-transports-and-authFri, 06 Nov 2026 00:00:00 GMTDeep divesBuild an MCP server on the 2026-07-28 spec: tools, resources, prompts, stdio vs Streamable HTTP, stateless requests, OAuth 2.1 and security rules. https://bydtech.in/blog/ai-agent-code-sandboxinghttps://bydtech.in/blog/ai-agent-code-sandboxingFri, 06 Nov 2026 00:00:00 GMTDeep divesHow to sandbox code an AI agent writes and runs: the threat model, containers, seccomp, gVisor, Firecracker microVMs, Kata, WebAssembly and practical limits. https://bydtech.in/blog/file-upload-system-designhttps://bydtech.in/blog/file-upload-system-designFri, 06 Nov 2026 00:00:00 GMTDeep divesHow to design file uploads: presigned URLs straight to object storage, multipart and resumable uploads, magic-number checks, malware scans and a CDN. https://bydtech.in/blog/zero-downtime-database-migrationshttps://bydtech.in/blog/zero-downtime-database-migrationsFri, 06 Nov 2026 00:00:00 GMTDeep divesHow to change a PostgreSQL schema without downtime: expand and contract, safe recipes for columns and indexes, batched backfills, lock_timeout and rollback. https://bydtech.in/blog/lora-fine-tuning-explainedhttps://bydtech.in/blog/lora-fine-tuning-explainedThu, 05 Nov 2026 00:00:00 GMTDeep divesWhen fine-tuning an LLM beats prompting or RAG, and how LoRA and QLoRA work: low-rank updates, rank and alpha, 4-bit training, adapter serving and data. https://bydtech.in/blog/llm-guardrails-architecturehttps://bydtech.in/blog/llm-guardrails-architectureThu, 05 Nov 2026 00:00:00 GMTDeep divesHow LLM guardrails fit together: input checks, output checks, tool-call policy, rules vs classifiers vs LLM judges, streaming checks and fail-closed design. https://bydtech.in/blog/text-to-sql-agentshttps://bydtech.in/blog/text-to-sql-agentsThu, 05 Nov 2026 00:00:00 GMTDeep divesHow text-to-SQL agents turn a question into SQL: schema retrieval, generation, validation and self-correction, and why the database must enforce the limits. https://bydtech.in/blog/search-autocomplete-system-designhttps://bydtech.in/blog/search-autocomplete-system-designThu, 05 Nov 2026 00:00:00 GMTDeep divesHow to design search autocomplete: a trie with top-k per node, prefix indexes, suggestions from query logs, edge caching, typo tolerance and the client. https://bydtech.in/blog/distributed-lockshttps://bydtech.in/blog/distributed-locksThu, 05 Nov 2026 00:00:00 GMTDeep divesHow distributed locks fail under pauses, clock drift and slow networks, why leases need fencing tokens, the Redlock debate, and the alternatives to a lock. https://bydtech.in/blog/feature-flags-system-designhttps://bydtech.in/blog/feature-flags-system-designThu, 05 Nov 2026 00:00:00 GMTDeep divesHow feature flag systems evaluate rules, bucket users for percentage rollouts, stream updates, fail safely and stay auditable, and how to clean up stale flags. https://bydtech.in/blog/ai-agent-memoryhttps://bydtech.in/blog/ai-agent-memoryWed, 04 Nov 2026 00:00:00 GMTDeep divesHow AI agent memory works: context windows, rolling summaries, retrieval by relevance, recency and importance, write policies, MemGPT, privacy and evaluation. https://bydtech.in/blog/multi-agent-orchestration-patternshttps://bydtech.in/blog/multi-agent-orchestration-patternsWed, 04 Nov 2026 00:00:00 GMTDeep divesMulti-agent orchestration patterns compared: pipelines, routing, orchestrator-workers, voting, evaluator loops, handoffs and debate, plus costs and failures. https://bydtech.in/blog/llm-inference-serving-batchinghttps://bydtech.in/blog/llm-inference-serving-batchingWed, 04 Nov 2026 00:00:00 GMTDeep divesHow LLM serving works: prefill and decode, TTFT and inter-token latency, continuous batching, chunked prefill, disaggregation and speculative decoding. https://bydtech.in/blog/event-sourcing-and-cqrshttps://bydtech.in/blog/event-sourcing-and-cqrsWed, 04 Nov 2026 00:00:00 GMTDeep divesHow event sourcing and CQRS work: event streams and replay, projections, snapshots, upcasting, idempotent read models, GDPR deletion and when to avoid them. https://bydtech.in/blog/consistency-models-cap-theoremhttps://bydtech.in/blog/consistency-models-cap-theoremWed, 04 Nov 2026 00:00:00 GMTDeep divesWhat the CAP theorem really proves, PACELC, linearizable, causal and eventual consistency, quorums where R + W > N, and how to pick a model for each feature. https://bydtech.in/blog/notification-system-designhttps://bydtech.in/blog/notification-system-designWed, 04 Nov 2026 00:00:00 GMTDeep divesHow to design notifications: preferences, templates, quiet hours by time zone, dedupe keys, per-channel queues, retries, push tokens, bounces and unsubscribe. https://bydtech.in/blog/site-agent-releasehttps://bydtech.in/blog/site-agent-releaseTue, 03 Nov 2026 00:00:00 GMTReleasesWe published site-agent, an open-source tag that lets visitors ask your site to do things and calls only the WebMCP tools your page declared. https://bydtech.in/blog/eu-ai-act-december-2-ai-content-markinghttps://bydtech.in/blog/eu-ai-act-december-2-ai-content-markingTue, 03 Nov 2026 00:00:00 GMTNewsArticle 50 of the EU AI Act has applied since 2 August 2026. Older generative AI systems have until 2 December 2026 to mark their output. A checklist. https://bydtech.in/blog/audit-tools-github-actionshttps://bydtech.in/blog/audit-tools-github-actionsTue, 03 Nov 2026 00:00:00 GMTReleasessite-audit, port-audit and agent-config-audit now ship as GitHub Actions on the Marketplace: one step scans, reports in the job summary and fails on Fix now. https://bydtech.in/blog/rag-explainedhttps://bydtech.in/blog/rag-explainedTue, 03 Nov 2026 00:00:00 GMTDeep divesHow retrieval-augmented generation works, step by step: parsing, chunking, embedding, indexing, retrieval, reranking and prompting, with a Python sketch. https://bydtech.in/blog/chunking-strategies-for-raghttps://bydtech.in/blog/chunking-strategies-for-ragTue, 03 Nov 2026 00:00:00 GMTDeep divesFixed-size, recursive, structure-aware, semantic, parent-child, contextual and late chunking for RAG: how each splits text, its trade-offs, and how to test it. https://bydtech.in/blog/hybrid-search-bm25-vector-searchhttps://bydtech.in/blog/hybrid-search-bm25-vector-searchTue, 03 Nov 2026 00:00:00 GMTDeep divesHow BM25 and vector search fail in different ways, how reciprocal rank fusion and weighted score fusion merge them, and how to tune and test the result. https://bydtech.in/blog/how-to-evaluate-raghttps://bydtech.in/blog/how-to-evaluate-ragTue, 03 Nov 2026 00:00:00 GMTDeep divesEvaluate RAG in two halves: retrieval with recall@k, MRR and nDCG, and answers with faithfulness and relevance, on a labeled test set that runs in CI. https://bydtech.in/blog/how-llm-tool-calling-workshttps://bydtech.in/blog/how-llm-tool-calling-worksTue, 03 Nov 2026 00:00:00 GMTDeep divesHow LLM tool calling works: tool definitions, what the model returns, the agent loop, parallel calls, tool_choice, strict schemas, errors, safety and MCP. https://bydtech.in/blog/ai-chat-assistant-system-designhttps://bydtech.in/blog/ai-chat-assistant-system-designTue, 03 Nov 2026 00:00:00 GMTDeep divesA system design for a website AI assistant: ingestion on publish, hybrid retrieval, SSE streaming, caching layers, guardrails, evaluation and failure modes. https://bydtech.in/blog/rate-limiter-system-designhttps://bydtech.in/blog/rate-limiter-system-designTue, 03 Nov 2026 00:00:00 GMTDeep divesHow fixed window, sliding window and token bucket rate limiters work, how to hold one limit across many servers, and how to answer with 429 and Retry-After. https://bydtech.in/blog/idempotency-and-job-queueshttps://bydtech.in/blog/idempotency-and-job-queuesTue, 03 Nov 2026 00:00:00 GMTDeep divesWhy queues deliver at least once, and how idempotency keys, dedupe tables, the outbox pattern and dead-letter queues give you exactly-once effects. https://bydtech.in/blog/caching-strategieshttps://bydtech.in/blog/caching-strategiesTue, 03 Nov 2026 00:00:00 GMTDeep divesCache-aside, read-through, write-through and write-back compared, with TTL jitter, invalidation, stampede protection, eviction and HTTP caching headers. https://bydtech.in/blog/database-indexing-explainedhttps://bydtech.in/blog/database-indexing-explainedTue, 03 Nov 2026 00:00:00 GMTDeep divesHow B-tree indexes work, what they cost, how to order composite indexes, when covering and partial indexes help, and how to read a PostgreSQL query plan. https://bydtech.in/blog/text-embeddings-explainedhttps://bydtech.in/blog/text-embeddings-explainedTue, 03 Nov 2026 00:00:00 GMTDeep divesHow text embeddings work: contrastive training, cosine vs dot product, Matryoshka dimensions, token limits, int8 and binary quantization, and model choice. https://bydtech.in/blog/prompt-caching-and-kv-cachehttps://bydtech.in/blog/prompt-caching-and-kv-cacheTue, 03 Nov 2026 00:00:00 GMTDeep divesHow the KV cache and prompt caching work: prefill and decode, prefix reuse in vLLM and SGLang, OpenAI, Anthropic and Gemini caching, and how to get hits. https://bydtech.in/blog/vector-search-ann-indexeshttps://bydtech.in/blog/vector-search-ann-indexesTue, 03 Nov 2026 00:00:00 GMTDeep divesHow approximate nearest neighbor indexes work: HNSW graphs, IVF clustering, product quantization, DiskANN, filtered search, and how to benchmark recall. https://bydtech.in/blog/reranking-cross-encodershttps://bydtech.in/blog/reranking-cross-encodersTue, 03 Nov 2026 00:00:00 GMTDeep divesHow rerankers work in search and RAG: bi-encoders vs cross-encoders, ColBERT late interaction, LLM rerankers, candidate depth, score thresholds and evaluation. https://bydtech.in/blog/constrained-decoding-structured-outputshttps://bydtech.in/blog/constrained-decoding-structured-outputsTue, 03 Nov 2026 00:00:00 GMTDeep divesHow LLM structured outputs work: logit masking, JSON Schema compiled to automata, token boundaries, provider strict-mode limits, quality and retries. https://bydtech.in/blog/llm-observability-tracinghttps://bydtech.in/blog/llm-observability-tracingTue, 03 Nov 2026 00:00:00 GMTDeep divesWhat to record on every LLM call, how to trace agent runs with the OpenTelemetry GenAI conventions, redaction, sampling, alerts and online evals. https://bydtech.in/blog/webhook-delivery-system-designhttps://bydtech.in/blog/webhook-delivery-system-designTue, 03 Nov 2026 00:00:00 GMTDeep divesHow to design webhook delivery: an outbox and a queue, retries with backoff and jitter, HMAC signatures, SSRF checks, dead letters, replay and delivery logs. https://bydtech.in/blog/consistent-hashing-and-shardinghttps://bydtech.in/blog/consistent-hashing-and-shardingTue, 03 Nov 2026 00:00:00 GMTDeep divesWhen to shard, how to pick a shard key, and how consistent hashing, virtual nodes, rendezvous and jump hash keep keys in place when servers are added.