Week 02 — Tool calling, API design, two pointers¶

At a glance

Phase 1: Foundations · Starts: Mon 05 Oct 2026 · Planned: 16h 45m

Build: Capstone: typed tool calling loop (runbook search, service status, ticket create)

By day¶

System Design · 45 min · API design: idempotency keys, cursor pagination, REST vs gRPC vs GraphQL trade-offs — decide for capstone's external API → API design: REST, gRPC, GraphQL, idempotency, pagination w02-sd-1
DSA · 25 min · Valid Sudoku (row/col/box sets) → Arrays & hashing · resource w02-dsa-1
Python · 20 min · Rep: typing.Protocol for an LLMClient port; structural vs nominal typing, runtime_checkable caveats → Advanced typing: generics, Protocols, ParamSpec, TypedDict · resource w02-py-1
Communication · 30 min · Grammar: Prepositions and collocations → Week 02 drills · day 1 w02-comm-1

Agentic AI · 60 min · Structured outputs deep dive: constrained decoding vs tool-forcing vs validate-and-retry; measure failure rate on 50 synthetic incident texts → Prompting & structured outputs w02-ai-1
DSA · 30 min · Valid Palindrome + Two Sum II - Input Array Is Sorted → Two pointers · resource w02-dsa-2
Communication · 30 min · Vocabulary: Word nuance: affect/impact/influence, ensure/assure/insure → Week 02 drills · day 2 w02-comm-2

DSA · 25 min · 3Sum (sort + two pointers, skip duplicates) → Two pointers · resource w02-dsa-3
Architecture · 45 min · Architecture styles: modular monolith vs microservices vs event-driven — score each against the capstone's characteristics → Architecture styles: modular monolith → microservices → event-driven → serverless · resource w02-arch-1
Python · 20 min · Rep: data model — eq/hash/repr on a value object; why dataclass(frozen=True) vs Pydantic here → Python data model & dunder protocols · resource w02-py-2
Communication · 30 min · Speaking drill: Word stress and sentence rhythm → Week 02 drills · day 3 w02-comm-3

Agentic AI · 60 min · LLM internals for engineers: KV cache, context window cost, prefill vs decode latency — estimate p50/p95 latency and cost per /triage call → LLM fundamentals: tokens, transformers, sampling, reasoning models · resource w02-ai-2
Java/Spring AI · 30 min · Spring AI 2.0 ChatClient fluent API: system/user prompts, options, streaming — hello-world against Ollama → Spring AI 2.0 fundamentals: ChatClient, advisors, structured output · resource w02-java-1
Communication · 30 min · Idioms & phrasal verbs: Progress and delay: behind schedule, move the needle → Week 02 drills · day 4 w02-comm-4

System Design · 45 min · Load balancing: L4 vs L7, algorithms, health checks — work through Sam Who's interactive essay → Load balancing & proxies · resource w02-sd-2
DSA · 25 min · Container With Most Water (greedy pointer move proof) → Two pointers · resource w02-dsa-4
Python · 20 min · Rep: Pydantic discriminated unions for tool results (Success | ToolError) with Field(discriminator=...) → Pydantic v2 in depth w02-py-3
Communication · 30 min · Writing: Clear status updates → Week 02 drills · day 5 w02-comm-5

Agentic AI · 120 min · Capstone build: 3 typed tools (search_runbooks stub, get_service_status, create_ticket) with Pydantic arg schemas, error-as-result returns and a hand-rolled tool loop with max-steps → Tool calling & function design w02-ai-3
DSA · 30 min · Trapping Rain Water (two pointers with left/right max) → Two pointers · resource w02-dsa-5
Java/Spring AI · 30 min · Spring AI structured output: .entity(IncidentTriage.class) with a Java record; compare failure behaviour to Python → Spring AI 2.0 fundamentals: ChatClient, advisors, structured output w02-java-2
Communication · 30 min · Record & shadow: Word stress and sentence rhythm → Week 02 drills · day 6 w02-comm-6

System Design · 45 min · Written design: estimation write-up — size the capstone for 50 teams, 2k incidents/day, 10 tool calls each (tokens, $/month, QPS, storage) → Interview framework & back-of-envelope estimation w02-sd-3
Architecture · 30 min · Write ADR 002 for capstone: modular monolith (Python) + separate Spring AI MCP service → Architecture styles: modular monolith → microservices → event-driven → serverless · resource w02-arch-2
Staff+ · 30 min · Artifact: archetype self-assessment — which archetype you target, evidence you have, 3 gaps (docs/log/staff/archetype.md) → Staff archetypes & operating at Staff+ · resource w02-staff-1
Communication · 30 min · Soft skills + weekly review: Giving feedback with SBI (situation-behaviour-impact) → Week 02 drills · day 7 w02-comm-7
Review · 60 min · Explain-it-back (record 5 min): tool-calling loop design and why tool errors are returned as results; 15 flashcards on API design → Tool calling & function design w02-rev-1