https://pratikdhanave.com/blog/ Long-form writing on multi-agent AI, medical AI governance, HIPAA-aware architecture, and cloud-native systems. en Sat, 10 Oct 2026 00:00:00 +0000 https://pratikdhanave.com/blog/posts/fineng-08-risk-management-and-models-in-practice.html https://pratikdhanave.com/blog/posts/fineng-08-risk-management-and-models-in-practice.html Sat, 10 Oct 2026 00:00:00 +0000 Risk management is the purpose every model served. Value at Risk and its tail-hiding flaw (so expected shortfall), fat tails, hedging with the Greeks, market microstructure as a systems problem, and the overriding lesson: models are approximations that will eventually be wrong — so respect the tails and stay humble. (Educational, not financial advice.) Financial Engineering Risk Management Value at Risk Model Risk Quantitative Finance Mathematical Finance https://pratikdhanave.com/blog/posts/llmops-08-operating-in-production.html https://pratikdhanave.com/blog/posts/llmops-08-operating-in-production.html Sat, 10 Oct 2026 00:00:00 +0000 Running an LLM app means treating cost and latency as first-class constraints (model routing, token hygiene, caching, streaming), preparing for incidents as certainties (provider failures, instant rollback, blameless postmortems), and the closing idea: operating is continuous improvement — a tightening loop, because an LLM app is never done. LLMOps Cost Optimization SRE Production AI Engineering LLMs Streaming https://pratikdhanave.com/blog/posts/fineng-07-randomness-stochastic-models-monte-carlo.html https://pratikdhanave.com/blog/posts/fineng-07-randomness-stochastic-models-monte-carlo.html Fri, 09 Oct 2026 00:00:00 +0000 Prices are random, so finance is a probabilistic discipline. Brownian motion and geometric Brownian motion model how a price wanders; Monte Carlo simulation turns that model into a universal engine for pricing and risk — the computational core of modern quant finance, and territory engineers find familiar. (Educational, not financial advice.) Financial Engineering Monte Carlo Stochastic Processes Simulation Quantitative Finance Mathematical Finance https://pratikdhanave.com/blog/posts/llmops-07-observability-and-tracing.html https://pratikdhanave.com/blog/posts/llmops-07-observability-and-tracing.html Fri, 09 Oct 2026 00:00:00 +0000 'The model messed up' is never a useful diagnosis. Why ordinary monitoring isn't enough, tracing as the spine (spans capturing the rendered prompt, retrieved context, tool calls, raw output, tokens/cost), and turning aggregated traces into metrics, quality trends, and drift detection. LLMOps Observability Tracing Production AI Engineering Reliability LLMs Prompt Engineering https://pratikdhanave.com/blog/posts/fineng-06-option-pricing-black-scholes.html https://pratikdhanave.com/blog/posts/fineng-06-option-pricing-black-scholes.html Thu, 08 Oct 2026 00:00:00 +0000 Black-Scholes turned option pricing into engineering: price an option by the cost of continuously replicating it, so no-arbitrage pins its value and the underlying's expected return drops out entirely. Risk-neutral valuation, what the model assumes, implied volatility, and the Greeks traders actually hedge. (Educational, not financial advice.) Financial Engineering Black-Scholes Options Derivatives Quantitative Finance Mathematical Finance https://pratikdhanave.com/blog/posts/llmops-06-guardrails-and-safety-in-production.html https://pratikdhanave.com/blog/posts/llmops-06-guardrails-and-safety-in-production.html Thu, 08 Oct 2026 00:00:00 +0000 An LLM can say anything, which is the problem. The new failure modes — hallucination, prompt injection, data leakage, harmful output — and layered, fail-closed guardrails (input, output, action) with humans in the loop, treated as part of the lifecycle and red-teamed. LLMOps AI Safety Guardrails Security AI Engineering Production LLMs Prompt Engineering https://pratikdhanave.com/blog/posts/fineng-05-derivatives-and-no-arbitrage.html https://pratikdhanave.com/blog/posts/fineng-05-derivatives-and-no-arbitrage.html Wed, 07 Oct 2026 00:00:00 +0000 A derivative's value derives from something else — and it's priced not by prediction but by the no-arbitrage principle: if a portfolio replicates a payoff, it must cost the same. Forwards, futures, options, payoff shapes, and the replication argument that prices a forward and links calls to puts. (Educational, not financial advice.) Financial Engineering Derivatives No-Arbitrage Options Quantitative Finance Mathematical Finance https://pratikdhanave.com/blog/posts/llmops-05-online-evaluation-and-feedback.html https://pratikdhanave.com/blog/posts/llmops-05-online-evaluation-and-feedback.html Wed, 07 Oct 2026 00:00:00 +0000 Offline evals say a change is safe to ship; online evals say whether it worked. The intrinsic offline–online gap, deploying to measure (canary, A/B, shadow), capturing explicit and implicit feedback, and the pipeline that turns production behavior into continuous improvement. LLMOps Evaluation A/B Testing Production AI Engineering Testing https://pratikdhanave.com/blog/posts/fineng-04-risk-and-return.html https://pratikdhanave.com/blog/posts/fineng-04-risk-and-return.html Tue, 06 Oct 2026 00:00:00 +0000 Higher expected return requires more risk — and the surprising twist is that diversification reduces risk for free when assets are imperfectly correlated. Volatility, covariance, the efficient frontier, systematic vs. idiosyncratic risk, and how CAPM prices risk through beta. (Educational, not financial advice.) Financial Engineering Portfolio Theory Risk CAPM Quantitative Finance Mathematical Finance https://pratikdhanave.com/blog/posts/llmops-04-eval-driven-development.html https://pratikdhanave.com/blog/posts/llmops-04-eval-driven-development.html Tue, 06 Oct 2026 00:00:00 +0000 The practice that separates improving an LLM app from merely changing it. The curated eval set as the core asset, how to score open-ended output (deterministic checks, reference metrics, LLM-as-judge, human review), and how to gate every change on evals and compare B vs A. LLMOps Evaluation AI Engineering Production LLMs https://pratikdhanave.com/blog/posts/fineng-03-bonds-and-the-yield-curve.html https://pratikdhanave.com/blog/posts/fineng-03-bonds-and-the-yield-curve.html Mon, 05 Oct 2026 00:00:00 +0000 A bond is the simplest real application of discounting: fixed future cash flows, priced by present value. From there come yield, the inverse price-yield relationship, duration and convexity as interest-rate risk, and the yield curve as the market's price of time. (Educational, not financial advice.) Financial Engineering Bonds Fixed Income Interest Rates Quantitative Finance Mathematical Finance https://pratikdhanave.com/blog/posts/llmops-03-prompt-and-version-management.html https://pratikdhanave.com/blog/posts/llmops-03-prompt-and-version-management.html Mon, 05 Oct 2026 00:00:00 +0000 In an LLM app the prompt is the source code — where behavior is defined and most bugs happen. Treat prompts like code: version the whole configuration (template + model + parameters + context strategy), decouple them from app code, and route every change through review, eval, rollout, and rollback. LLMOps Prompt Engineering Version Control Production AI Engineering LLMs Evaluation https://pratikdhanave.com/blog/posts/fd-stack-04-building-a-forward-deployment-function.html https://pratikdhanave.com/blog/posts/fd-stack-04-building-a-forward-deployment-function.html Sun, 04 Oct 2026 00:00:00 +0000 A practical blueprint for founders: when to start forward deployment, who to hire first, how to scope engagements, what to measure, and how to keep it from eating your margins — plus the four failure modes (agency, hero, science-project, over-serving) and a 90-day start. Forward Deployed Engineer Go-to-Market AI Engineering Product https://pratikdhanave.com/blog/posts/fineng-02-the-time-value-of-money.html https://pratikdhanave.com/blog/posts/fineng-02-the-time-value-of-money.html Sun, 04 Oct 2026 00:00:00 +0000 A dollar today is worth more than a dollar tomorrow — and that single fact, made precise, is how everything in finance gets valued. Present value, future value, discounting, and the discounted-cash-flow method that prices anything with a future payoff. (Educational, not financial advice.) Financial Engineering Time Value of Money Discounting Valuation Quantitative Finance Mathematical Finance https://pratikdhanave.com/blog/posts/llmops-02-the-llmops-lifecycle.html https://pratikdhanave.com/blog/posts/llmops-02-the-llmops-lifecycle.html Sun, 04 Oct 2026 00:00:00 +0000 The LLMOps lifecycle — develop, evaluate, deploy, observe, improve — as a loop an LLM app lives inside, not a line it runs once. Why evaluation is a first-class stage appearing twice, why production feedback flows backward, and why the loop turns whether you act or not. LLMOps Lifecycle AI Engineering Production LLMs Evaluation https://pratikdhanave.com/blog/posts/fd-stack-03-the-forward-deployed-leader.html https://pratikdhanave.com/blog/posts/fd-stack-03-the-forward-deployed-leader.html Sat, 03 Oct 2026 00:00:00 +0000 Forward deployment that isn't run as a disciplined function becomes an unscalable custom-dev shop. The Forward Deployed Leader owns the motion as a business — team design, scoping discipline, unit economics, and the decision of what becomes product — and names the invisible trap: it feels like success right up until the margins invert. Forward Deployed Engineer Go-to-Market Leadership Strategy https://pratikdhanave.com/blog/posts/fineng-01-what-financial-engineering-is.html https://pratikdhanave.com/blog/posts/fineng-01-what-financial-engineering-is.html Sat, 03 Oct 2026 00:00:00 +0000 Financial engineering applies math, statistics, and computation to money — and under the jargon it rests on three ideas an engineer will recognize: the time value of money, the trade-off between risk and return, and no-arbitrage. This series builds quant finance from those foundations. (Educational, not financial advice.) Financial Engineering Quantitative Finance Mathematical Finance Fundamentals https://pratikdhanave.com/blog/posts/llmops-01-what-llmops-is.html https://pratikdhanave.com/blog/posts/llmops-01-what-llmops-is.html Sat, 03 Oct 2026 00:00:00 +0000 Getting an LLM to work in a demo is easy; keeping it working, safe, and affordable for a year is a discipline. What LLMOps is, how it differs from MLOps, and the five properties of LLMs — opaque external models, prompt-defined behavior, fuzzy correctness, non-determinism, new failures/costs — that force a different playbook. LLMOps MLOps AI Engineering Production LLMs Prompt Engineering https://pratikdhanave.com/blog/posts/fd-stack-02-the-forward-deployed-architect.html https://pratikdhanave.com/blog/posts/fd-stack-02-the-forward-deployed-architect.html Fri, 02 Oct 2026 00:00:00 +0000 The forward deployed engineer builds a win inside one customer; the Forward Deployed Architect makes it survive the next ten — reference architecture, security and governance that passes review, and the reuse decision that separates product from one-off. The missing layer between heroics and product. Forward Deployed Engineer AI Architecture Solution Architecture AI Engineering https://pratikdhanave.com/blog/posts/gpueng-08-profiling-and-the-optimization-workflow.html https://pratikdhanave.com/blog/posts/gpueng-08-profiling-and-the-optimization-workflow.html Fri, 02 Oct 2026 00:00:00 +0000 You can't optimize what you haven't measured. The optimization loop — baseline, profile, find the dominant limit, apply the matching lever, repeat — plus knowing when to stop, leaning on the ecosystem, and the durable mental model behind it all. GPU Profiling Performance Inference https://pratikdhanave.com/blog/posts/fd-stack-01-the-forward-deployment-stack.html https://pratikdhanave.com/blog/posts/fd-stack-01-the-forward-deployment-stack.html Thu, 01 Oct 2026 00:00:00 +0000 The forward deployed engineer gets all the attention, but forward deployment is bigger than one role — it's a go-to-market operating model with three layers: build (engineer), scale (architect), run (leader). Seeing the whole stack is how AI companies turn embedded engineering from a heroic cost center into a repeatable engine. Forward Deployed Engineer Go-to-Market AI Engineering Strategy https://pratikdhanave.com/blog/posts/gpueng-07-memory-management-kv-cache.html https://pratikdhanave.com/blog/posts/gpueng-07-memory-management-kv-cache.html Thu, 01 Oct 2026 00:00:00 +0000 The KV cache often exceeds the model weights in HBM, and naive contiguous allocation wrecks concurrency with fragmentation. PagedAttention applies operating-system paging to fix it — a core inference bottleneck solved by borrowing a classic systems idea. GPU KV Cache PagedAttention Inference Performance LLMs Concurrency https://pratikdhanave.com/blog/posts/gpueng-06-the-attention-bottleneck-flashattention.html https://pratikdhanave.com/blog/posts/gpueng-06-the-attention-bottleneck-flashattention.html Wed, 30 Sep 2026 00:00:00 +0000 Naive attention materializes an N×N matrix in HBM — memory-bound and quadratic. FlashAttention fixes it with pure roofline reasoning: tile into shared memory, use an online softmax, fuse the kernel, never touch HBM. The archetype of all inference optimization. GPU FlashAttention Transformers Inference Performance https://pratikdhanave.com/blog/posts/gpueng-05-numeric-precision.html https://pratikdhanave.com/blog/posts/gpueng-05-numeric-precision.html Tue, 29 Sep 2026 00:00:00 +0000 FP32, FP16, BF16, and INT8/INT4 — how a float splits bits between range and precision, why bf16 won, and why fewer bits is a double win: less data to stream from HBM and faster tensor-core math. Quantization and why LLMs tolerate it. GPU Quantization Precision Inference Performance LLMs https://pratikdhanave.com/blog/posts/gpueng-04-the-cuda-execution-model.html https://pratikdhanave.com/blog/posts/gpueng-04-the-cuda-execution-model.html Mon, 28 Sep 2026 00:00:00 +0000 Threads, warps, blocks, and grids; SIMT and branch divergence; and the cleverest trick — hiding slow memory by oversubscribing warps (occupancy). Why a GPU can be underutilized without being either compute- or memory-bound. GPU CUDA Performance Inference https://pratikdhanave.com/blog/posts/ai-fde-08-from-bespoke-ai-to-product.html https://pratikdhanave.com/blog/posts/ai-fde-08-from-bespoke-ai-to-product.html Mon, 28 Sep 2026 00:00:00 +0000 Every AI FDE builds one-offs — a bespoke deployment for one customer's data, workflow, and trust. The ones who create lasting value turn those one-offs into product: the patterns that repeat become a platform, the platform makes the next deployment faster, and field learnings flow back to shape what gets built. The flywheel that turns bespoke AI work into a compounding asset — and the AI FDE career arc. Forward Deployed Engineer AI Engineering Product Strategy LLMs https://pratikdhanave.com/blog/posts/gpueng-03-the-roofline-model.html https://pratikdhanave.com/blog/posts/gpueng-03-the-roofline-model.html Sun, 27 Sep 2026 00:00:00 +0000 The roofline model answers the one question that governs every optimization — compute-bound or memory-bound? — from arithmetic intensity (FLOPs per byte). Why LLM prefill is compute-bound and decode is memory-bound, and why batching is the master throughput lever. GPU Roofline Performance Inference LLMs https://pratikdhanave.com/blog/posts/ai-fde-07-productionizing-and-handover.html https://pratikdhanave.com/blog/posts/ai-fde-07-productionizing-and-handover.html Sun, 27 Sep 2026 00:00:00 +0000 A pilot that works is not a system the customer can run. Productionizing AI means making it reliable, affordable, fast, and observable enough to be real infrastructure — then handing it over so the customer operates it without you. The FDE's goal, in the end, is to make themselves unnecessary. Cost, latency, drift, observability, and a handover that transfers the eval discipline, not just the code. Forward Deployed Engineer MLOps AI Engineering Production LLMs Evaluation Observability https://pratikdhanave.com/blog/posts/gpueng-02-the-memory-hierarchy.html https://pratikdhanave.com/blog/posts/gpueng-02-the-memory-hierarchy.html Sat, 26 Sep 2026 00:00:00 +0000 The GPU memory hierarchy from registers to HBM, the difference between capacity and bandwidth limits, and the defining fact of LLM generation: it's memory-bandwidth-bound because every token reads the whole model from HBM — which is why batching and quantization work. GPU Memory Bandwidth Inference Performance LLMs https://pratikdhanave.com/blog/posts/ai-fde-06-integrating-ai-into-workflows.html https://pratikdhanave.com/blog/posts/ai-fde-06-integrating-ai-into-workflows.html Sat, 26 Sep 2026 00:00:00 +0000 A technically excellent AI system that nobody uses has delivered zero value. The last mile of an AI deployment is not the model — it's fitting the system into how real people actually do their jobs, designing an interface that handles uncertainty honestly, and managing the human change of introducing AI into someone's work. Human-in-the-loop, calibrated reliance, autonomy levels, and change management. Forward Deployed Engineer AI Engineering UX Human-in-the-Loop LLMs https://pratikdhanave.com/blog/posts/recsys-08-production-recommenders.html https://pratikdhanave.com/blog/posts/recsys-08-production-recommenders.html Sat, 26 Sep 2026 00:00:00 +0000 Building a good model is half the work; running a recommender in production is the other half. Real systems must serve in milliseconds, stay fresh as the catalog and tastes change, handle cold start gracefully, resist the filter bubbles their own optimization creates, and be monitored like any critical service. This closing post assembles everything — and gives a practical path to building one. Recommender Systems Production AI Engineering Machine Learning Personalization https://pratikdhanave.com/blog/posts/gpueng-01-why-gpus-power-ai.html https://pratikdhanave.com/blog/posts/gpueng-01-why-gpus-power-ai.html Fri, 25 Sep 2026 00:00:00 +0000 Every LLM runs on a GPU, but the spec-sheet FLOPS number is a near-fiction in practice. Why neural nets map onto GPUs (matrix multiply, SMs, tensor cores), and the governing intuition of the series: compute is cheap, data movement is expensive. GPU Inference CUDA Performance LLMs https://pratikdhanave.com/blog/posts/ai-fde-05-evaluation-and-trust.html https://pratikdhanave.com/blog/posts/ai-fde-05-evaluation-and-trust.html Fri, 25 Sep 2026 00:00:00 +0000 You cannot ship AI you cannot measure, and no enterprise grants a probabilistic system authority over real work on faith. Both problems have the same answer: evaluation. Building the customer's own eval set — real examples, their definition of correct — turns 'it seemed good in the demo' into a reliability number, and that number is how trust gets earned. Offline/online eval, safe failure, and trust. Forward Deployed Engineer AI Evaluation AI Engineering Trust LLMs Evaluation https://pratikdhanave.com/blog/posts/recsys-07-evaluating-recommenders.html https://pratikdhanave.com/blog/posts/recsys-07-evaluating-recommenders.html Fri, 25 Sep 2026 00:00:00 +0000 A recommender that scores well offline can flop in production, and a metric that looks like success can quietly harm the product. Evaluation is genuinely hard: offline metrics only approximate real behavior, the only ground truth is a live A/B test, and the very act of recommending shapes the data you learn from next. Offline metrics, online testing, the offline-online gap, and feedback loops. Recommender Systems Evaluation A/B Testing Machine Learning Personalization Testing https://pratikdhanave.com/blog/posts/onto-08-operating-knowledge-graphs-in-practice.html https://pratikdhanave.com/blog/posts/onto-08-operating-knowledge-graphs-in-practice.html Thu, 24 Sep 2026 00:00:00 +0000 A knowledge graph is a system you run, not a project you finish. Governing the ontology as it evolves, keeping the populated graph correct (freshness, re-resolution, provenance), and the honest question: given the cost, when is a knowledge graph the right tool — and when is a table or plain vector RAG cheaper and enough? Knowledge Graphs Data Governance Ontology Operations Semantic Web RAG https://pratikdhanave.com/blog/posts/ai-fde-04-grounding-ai-in-customer-data.html https://pratikdhanave.com/blog/posts/ai-fde-04-grounding-ai-in-customer-data.html Thu, 24 Sep 2026 00:00:00 +0000 A frontier model knows the public internet and nothing about the customer. All the value of an AI deployment comes from the opposite: making the model reason over the customer's own documents, records, and knowledge. Grounding — retrieval-augmented generation over messy, permissioned, incomplete data — is the technical heart of the AI FDE's job. With an interactive reference-architecture diagram. Forward Deployed Engineer RAG AI Engineering AI Architecture LLMs https://pratikdhanave.com/blog/posts/recsys-06-deep-learning-recommenders.html https://pratikdhanave.com/blog/posts/recsys-06-deep-learning-recommenders.html Thu, 24 Sep 2026 00:00:00 +0000 Modern large-scale recommenders are built on neural networks — not replacing the core ideas (embeddings, two stages) but supercharging them. This post covers the two workhorses: two-tower models for retrieval (the neural evolution of matrix factorization, built for fast ANN search) and rich neural ranking models that score the shortlist with cross-features, sequences, and multiple objectives. Recommender Systems Deep Learning Machine Learning AI Engineering Personalization Embeddings https://pratikdhanave.com/blog/posts/onto-07-the-semantic-layer.html https://pratikdhanave.com/blog/posts/onto-07-the-semantic-layer.html Wed, 23 Sep 2026 00:00:00 +0000 The enterprise ontology made operational: a governed layer that defines the business's entities, relationships, and metrics once, so dashboards, analysts, and AI agents all inherit the same definitions. Why 'revenue' means three things, and why AI made the semantic layer urgent — an agent is only as correct as its grasp of your terms. Knowledge Graphs Semantic Layer Data Engineering Ontology Semantic Web https://pratikdhanave.com/blog/posts/ai-fde-03-from-demo-to-pilot.html https://pratikdhanave.com/blog/posts/ai-fde-03-from-demo-to-pilot.html Wed, 23 Sep 2026 00:00:00 +0000 An AI demo is the easiest impressive thing to build and the most misleading: it runs on cherry-picked inputs, in a clean environment, with the failures edited out — and convinces everyone the problem is nearly solved when the real work has barely begun. Use the demo to win belief, then walk the customer honestly across the chasm to a pilot that survives real data. Forward Deployed Engineer AI Engineering Prototyping LLMs https://pratikdhanave.com/blog/posts/recsys-05-two-stage-architecture.html https://pratikdhanave.com/blog/posts/recsys-05-two-stage-architecture.html Wed, 23 Sep 2026 00:00:00 +0000 You can't run your best, most expensive model on millions of items per request. The near-universal answer is to split recommendation into two stages: a cheap, fast net that narrows millions to a few hundred, then a precise, heavier model that ranks those few. This retrieve-then-rank structure is the single most important architectural pattern in production recommenders — with an interactive pipeline diagram. Recommender Systems AI Architecture Machine Learning Systems Design Personalization https://pratikdhanave.com/blog/posts/onto-06-graphrag-ontology-grounded-retrieval.html https://pratikdhanave.com/blog/posts/onto-06-graphrag-ontology-grounded-retrieval.html Tue, 22 Sep 2026 00:00:00 +0000 The post the series built toward. Plain vector RAG retrieves disconnected chunks and misses relationships; GraphRAG retrieves a subgraph by traversing the knowledge graph, so multi-hop answers come from walked edges, not guessed connections — typed, verifiable, citable. When to use it, and why hybrid often wins. Knowledge Graphs GraphRAG RAG LLMs Ontology Semantic Web https://pratikdhanave.com/blog/posts/ai-fde-02-scoping-an-ai-use-case.html https://pratikdhanave.com/blog/posts/ai-fde-02-scoping-an-ai-use-case.html Tue, 22 Sep 2026 00:00:00 +0000 The most important decision an AI FDE makes happens before any code: which problem to point the model at. Choose a problem AI is genuinely suited for, with real value and a clear way to measure it, and the engagement can succeed. Choose AI theater — impressive-sounding but ill-fit — and no engineering saves it. Fit vs value, resisting AI theater, augmentation over automation, and picking the wedge. Forward Deployed Engineer AI Engineering Product LLMs https://pratikdhanave.com/blog/posts/recsys-04-matrix-factorization-and-embeddings.html https://pratikdhanave.com/blog/posts/recsys-04-matrix-factorization-and-embeddings.html Tue, 22 Sep 2026 00:00:00 +0000 Represent every user and item as a short vector of latent factors, and a user's affinity for an item becomes just the dot product of their vectors. Matrix factorization turned recommendation into learning good embeddings — the technique that won the Netflix Prize and the conceptual bridge from classical collaborative filtering to today's deep-learning systems and vector search. Recommender Systems Matrix Factorization Embeddings Machine Learning Personalization Vector Databases https://pratikdhanave.com/blog/posts/ebpf-08-ecosystem-and-building.html https://pratikdhanave.com/blog/posts/ebpf-08-ecosystem-and-building.html Tue, 22 Sep 2026 00:00:00 +0000 You rarely write raw eBPF bytecode by hand — a rich ecosystem of libraries, languages, and platforms sits on top, and one breakthrough (CO-RE, Compile Once Run Everywhere) solved the portability problem that once made eBPF programs fragile across kernel versions. This closing post surveys how you actually build with eBPF, and where it's heading. eBPF Linux Systems Programming Tooling https://pratikdhanave.com/blog/posts/onto-05-building-knowledge-graphs-entity-resolution.html https://pratikdhanave.com/blog/posts/onto-05-building-knowledge-graphs-entity-resolution.html Mon, 21 Sep 2026 00:00:00 +0000 The hard part isn't the ontology or the queries — it's construction. The pipeline from heterogeneous sources, and its make-or-break challenge: entity resolution (deciding when two records are the same thing). Under-merging fragments; over-merging asserts false connections; connectivity amplifies every error. Knowledge Graphs Entity Resolution Data Engineering Data Quality Ontology Semantic Web https://pratikdhanave.com/blog/posts/ai-fde-01-the-rise-of-the-ai-fde.html https://pratikdhanave.com/blog/posts/ai-fde-01-the-rise-of-the-ai-fde.html Mon, 21 Sep 2026 00:00:00 +0000 The forward deployed engineer was born at Palantir to bridge powerful software and messy customer reality. In the AI era the role exploded, because frontier models widened that gap: a model that dazzles in a demo is a long way from a system that works inside one company's data, workflows, and trust constraints. This series is the technical playbook for the engineer who closes that gap. Forward Deployed Engineer AI Engineering Career LLMs https://pratikdhanave.com/blog/posts/recsys-03-content-based-and-cold-start.html https://pratikdhanave.com/blog/posts/recsys-03-content-based-and-cold-start.html Mon, 21 Sep 2026 00:00:00 +0000 Collaborative filtering has one crippling blind spot: it knows nothing about brand-new users or items. Content-based filtering fills that gap by recommending based on what items are rather than who interacted with them. Understanding the cold-start problem — and how each approach handles it — is key to building a recommender that works from day one, which is why most real systems are hybrids. Recommender Systems Content-Based Filtering Machine Learning Cold Start Personalization https://pratikdhanave.com/blog/posts/ebpf-07-security.html https://pratikdhanave.com/blog/posts/ebpf-07-security.html Mon, 21 Sep 2026 00:00:00 +0000 Security is eBPF's most natural fit: the kernel sees every syscall, process, file access, and network connection, so an eBPF program in the kernel is perfectly positioned to watch for and stop malicious behavior in real time. This is why modern runtime-security tools — detecting and blocking threats on live systems — are increasingly built on eBPF. eBPF Security Linux DevSecOps Systems Programming https://pratikdhanave.com/blog/posts/voice-08-building-and-productionizing.html https://pratikdhanave.com/blog/posts/voice-08-building-and-productionizing.html Mon, 21 Sep 2026 00:00:00 +0000 A voice demo that works in a quiet room with a good headset is a long way from a voice agent that survives a noisy phone call from a real customer. Production voice AI has to handle bad audio, unpredictable humans, failures at every stage, and the peculiar demands of telephony — and be evaluated in ways text systems never require. Making a voice agent real. Voice AI Production Telephony AI Engineering Conversational AI Speech Evaluation https://pratikdhanave.com/blog/posts/onto-04-querying-with-sparql.html https://pratikdhanave.com/blog/posts/onto-04-querying-with-sparql.html Sun, 20 Sep 2026 00:00:00 +0000 Querying a graph is pattern matching, not joining: you describe a shape of nodes and edges and the engine finds every match. SPARQL (and Cypher for property graphs), path expressions for arbitrary-depth connection that SQL can't express, and inference-aware queries over the reasoned graph. Knowledge Graphs SPARQL Query Languages Semantic Web Ontology Databases