dhaga. blog
Deep engineering dives on the hardest problems we solved, and practical guides to how Dhaga fits real estate, finance, sales, recruiting and more.
Deep dives on the hardest problems we solved building Dhaga — each opens with a plain-language summary, then drops into the real engineering, with links to the exact code that shipped.
19 postsHow Dhaga fits specific lines of work — the daily relationship problems each one faces, and how capture, the knowledge graph, natural-language recall, and proactive intelligence map onto them.
10 postsPractical, step-by-step guides to networking and running a personal CRM — how to capture people, keep relationships warm, follow up on time, and turn scattered notes into a knowledge graph you can actually query.
21 postsThe story behind Dhaga and the ideas that shape it — why an AI-native personal CRM, and what we believe about who should own their relationship data.
4 postsTwo weeks, many parallel AI agents, one merge. Which layer of verification caught which bug, the boot deadlock our own tests could not see, and what we changed.
Notion- and Obsidian-style pages, databases, a daily Journal and a canvas, next to your people, meetings and follow-ups. Private by default, yours to export.
What India's DPDP Act asks of a signup checkbox, how GDPR and LGPD compare, and how we built a separate, recorded consent box, including what it cannot prove.
Delivering every minute didn't add a double-send; it made an old one routine. Why the queue is materialised, and why the claim must happen in the database.
Your daily summary now arrives at a time you choose, and nudges land before the meeting. Plus search on every table and notes while you add a person.
Twenty-nine minutes of CI cut to about ten by splitting the job and sharding the suite — but the changes that mattered most stopped a green tick lying.
A captcha on the signup form guards one door of five: in most auth libraries, magic links, email OTP, One Tap and OAuth all create accounts by default.
Widening row-level security with an OR so colleagues can share records exposes every column and costs your hottest index. Publish a projection instead.
On a light ground, a colour's fill and text contrast multiply to a constant — ours is 16.78 — so no one value clears 4.5:1 at both. Hence accent pairs.
Duplicate detection merged strangers via a tag suggester that groups by surname on purpose. Why types couldn't catch it, nor a smarter helper fix it.
Card scans came back confident, well-formed and wrong. The same 408px image read perfectly on a stronger model — the fix was the model, not more pixels.
We summarised 847 commits with LLMs and audited every line. No invented features — but true-when-written claims, and a 'shipped' path no user could reach.
An AI that tells you things about your friends is worthless unless you can check it. Every fact Dhaga derives carries a receipt, and dies with its note.
Covve vs Dhaga — a world-class card scanner built for trade-show teams, versus an AI-native personal CRM that turns your notes into a private graph.
folk vs Dhaga, honestly: folk shows which teammate knows a prospect; Dhaga shows what you know about them. Pricing, export and graph depth compared.
Mesh (formerly Clay) vs Dhaga: Mesh maps who is in your network; Dhaga records what you know and what you promised. Pricing, capture and provenance.
Orvo vs Dhaga — career relationship intelligence with a hand-drawn Network Map, versus a personal CRM whose graph is derived from the notes you write.
A personal CRM comparison for 2026 — Dhaga, Dex, folk, Mesh, Covve, Orvo and Monica on capture, AI, relationship graphs, platforms, privacy and price.
One message, one place — a rule that lived only in the prompt, checked by no code. Why a Zod refinement was the wrong lever, and why we repair bad plans.
A capture came back four fifths shorter, silently. Not a token limit, not summarisation — the model filtered to fit the schema. A schema is a filter too.
Twenty and Dhaga look alike on a checklist but solve different jobs. Compare capture, pipelines, AI, privacy, teams, and what feature parity would take.
Six sources under one Promise.all, a pool of three: HTTP 500 with one user. Why Promise.all multiplies concurrency when every read takes a connection.
Turns notes and cards into a searchable network — the best AI networking tools in India for 2026.
The best personal CRM apps in India for 2026 — privacy-first, AI, and free options compared so you can choose in five minutes.