Role / Pieces for Developers · Full-time

Five systems,
one person, 18 weeks.

I led marketing data infrastructure and marketing AI at Pieces, a developer-tools company. These five systems are what shipped, each one still running in production after the role ended.

0systems shipped, all still in production
0weeks, one person
0engineer owning data + AI
0Series A the team had raised
Full-timeGCP · Cloud RunBigQueryApache BeamRAGSeries A · $13.5M
01 · Overview

Five problems the growth team kept hitting.

Each system started with a problem no tool on the market solved for our use case, so I built it. Memory, listening, attribution, signal, and publishing. The summaries below tell the story end to end; each links to the full case study if you want the depth.

02 · Remembers
Company knowledge system

Osiris.

Ask what did we decide, what shipped last week, what are customers complaining about. Get a sourced answer in seconds. No more 15-minute hunt through chat, docs, and meeting notes.

to find one answer~15 min
→
sourced answer2.8s

A retrieval system (RAG) over 50+ chat spaces, daily transcripts, GitHub, and product docs on GCP. Every retrieved doc is graded before use, so answers don't mix last quarter's strategy with this quarter's pivot. The same knowledge base feeds the outbound and content systems below.

Full case study: Osiris →
03 · Listens
Social listening platform

Nexus.

Finds the conversations worth joining. The dev quietly evaluating you. The comparison thread. The buying-intent question. Scored and dropped in Slack. The noise gets filtered out.

found by hand0–2 / day
→
real ones, scored12+ / run

A single tweet means nothing alone. The system walks the full reply chain, enriches the profiles involved, then scores the whole context with AI. Runs on a schedule from the cloud without getting IP-banned, backs off when rate-limited, falls back to sequential if parallel fails.

Full case study: Nexus →
04 · Judges
Attribution + experimentation

Data Platform.

Answers the Monday question every growth lead has: which channels bring customers who actually stick, not just who sign up. One source of truth, pulled from tools that never talked to each other.

that never talked5+ tools
→
for the answer1 query

The whole measurement stack, from scratch: GTM tag management and event taxonomy, Apache Beam pipelines pulling web, social, and product-signup data into BigQuery, a lifecycle model joining session to feature usage to retention with multi-touch attribution. Idempotent pipelines retry without duplicates.

Full case study: Data Platform →
05 · Observes
Developer signal intelligence

Signal Intel.

Not stargazers. The people actually running your product in production, or shopping your competitor today. Each one researched, scored, and handed to reps as a draft. They review and send.

research / lead~15 min
→
evidence attached~5 min

Signals from 9 platforms (GitHub, Docker, PyPI, Stack Overflow, job boards, competitor comparisons, and more) linked into one identity per person, scored for real production evidence. Replaces 3–4 tools (Apollo, Clearbit, reo.dev, Sales Navigator) and refuses to draft for low-confidence matches.

Full case study: Signal Intel →
06 · Speaks
Multi-platform publishing

Content Engine.

One blog post becomes native posts for Dev.to, LinkedIn, X, and Medium, written to each platform's rules, quality-scored and deduped. Seconds, not the 2+ hours it takes by hand.

repurposing by hand2+ hrs
→
every platform12s

Every platform has different rules. The system scores each variant on a 100-point rubric and refuses to publish below 70. If two variants are more than 85% similar, it regenerates on its own. No daily babysitting.

Full case study: Content Engine →
07 · Reference
Tsavo Knott

“Nikhil led both marketing data infrastructure and marketing AI systems. He built Osiris, a marketing RAG system on GCP that indexed internal docs, GitHub content, and internal conversations into a queryable assistant. He developed custom agents for social listening, content workflows, and outbound personalization. He also owned core data plumbing end-to-end: GA4 integrations, GTM setup, and Apache Beam to BigQuery pipelines.”

Tsavo Knott · CEO & Technical Co-Founder, Pieces for Developers · LinkedIn ↗

Five systems. Still one person. Still running.

← Back to the systems