Weekly Tech Roundup — SpaceX/xAI merger, AI video, LLM wars
This week in tech was absolutely wild. SpaceX and xAI are merging into a $1.25 TRILLION entity. Elon's building the most vertically integrated AI-space company in history.
Read essayGGN · INDEX / 001
Accepting 2 projects
v4.2
Automations that think,
systems that ship.
I'm Gagan — a one-person studio building custom AI automations, agents, and data pipelines for teams that would rather do the work than manage the tools. Also a long-running native iOS shop, shipping Swift and SwiftUI apps to the App Store.
LIVE WORKFLOW · node_graph.v2⌘drag to pan
A narrow studio for a wide problem.
Every engagement is a custom build. These are the shapes they usually take.
S/01
Custom AI workflows
End-to-end automations across n8n, Zapier, and Make. Multi-step, multi-system, resilient under real traffic.
n8nZapierMakeTemporal
S/02
GPT & Claude agents
Reasoning agents with tool use, memory, and guardrails. Built with the model that fits the job — not the hype.
OpenAIAnthropicLangGraphAssistants
S/03
RAG & knowledge systems
Company-specific retrieval over your docs, tickets, and tribal knowledge. Ranked, cited, and actually accurate.
PineconepgvectorRagieLlamaIndex
S/04
Data pipelines & integrations
The plumbing behind everything else. ETL, webhooks, and event streams connecting your stack without duct tape.
AirbytedbtSegmentPostgres
S/05
Chatbots & support automation
Tier-1 deflection that customers actually like. Hand-off to humans when confidence dips, always with context.
IntercomZendeskFrontVoiceflow
S/06
Sales & CRM automation
Lead enrichment, routing, outbound sequencing, and revenue ops workflows built around your ICP.
HubSpotSalesforceClayApollo
S/07
Content generation systems
Structured pipelines that produce on-brand content at scale — with humans in the loop where they should be.
ClaudeRunwayIdeogramElevenLabs
S/08
Strategy & consulting
One-off audits, automation roadmaps, and build-vs-buy reviews. A few hours that often save months.
AuditRoadmapReviewRetainer
S/09
Native iOS apps
Swift and SwiftUI apps shipped to the App Store. Built for performance, accessibility, and the Apple ecosystem from the ground up — no cross-platform shortcuts.
SwiftSwiftUIXcodeObjective-C
How an engagement actually goes.
01 · DISCOVER
Diagnose, not prescribe.
A working session to map every manual step in the workflow. I leave with a list of candidates ranked by time-saved-per-dollar.
02 · DESIGN
Blueprint on paper.
A node-graph of your target system with every handoff, failure mode, and data contract named. You approve before anything runs.
03 · BUILD
Ship in slices.
Narrow vertical releases every 3–5 days. You see the system working on your data before it's fully finished.
04 · OPERATE
Own the outcome.
Monitoring, alerts, and iteration after launch. A monthly retainer keeps the system sharp as your business shifts.
A handful of recent builds.
A support agent that resolved 62% of tier-1 tickets at a Series B SaaS.
RAG over 40k help-center articles and historical tickets, routed through a Claude agent with a strict escalation policy and deflection scorecard.
62%Auto-resolved · 30-day average
Lead enrichment pipeline that tripled SDR throughput.
Clay + GPT-4 + HubSpot. Inbound forms scored, researched, and booked in under 90 seconds — no SDR involvement.
3.4×Meetings booked per SDR
A content system producing 120 localized assets per week.
Structured generation with brand-voice guardrails, human review queue, and Figma handoff. Replaced a four-person content pod.
18hSaved per week · ops team
Contract triage with cited reasoning.
Ingest any PDF, extract obligations, flag risk clauses against a firm-specific playbook. Citations back to the source paragraph.
92%Accuracy vs. senior associate
Operations co-pilot for a 20-person ops team.
A Slack-native agent with access to 11 internal tools. Fuzzy requests resolve into structured actions with a confirmation step.
$410kAnnualized labor saved
Native iOS app for a marketplace that crossed 1M installs.
Swift + SwiftUI rebuild of a legacy hybrid app. Shaved 60% off cold-start time, restored buttery scroll on long lists, and shipped iPad layouts in the same release.
1M+App Store installs · 12 months
On-device ML trust scoring for a fintech iOS app.
Core ML model running locally for fraud signals — no PII leaves the device. Shipped with biometric auth, App Store privacy nutrition labels, and a clean Objective-C ↔ Swift interop layer.
0msRound-trip · scoring stays on device
A live workflow, start to finish.
A real agent handling an inbound lead — trigger fires, agent reasons, actions dispatch to HubSpot, Notion, and Linear. The kind of thing that used to take a person eleven minutes.
◉ 01 · Trigger
◉ 02 · Reason
◉ 03 · Dispatch
◉ 04 · Observe
0+
Systems shipped
0k
Hours saved · client side
$0.0M
Labor value automated
0+
Products shipped · across AI & iOS
Words from people I ship for.
“I've worked closely with Gagan, and his ability to turn complex AI concepts into real, working products is exceptional. He doesn't just understand the tech — he knows how to execute at scale.”
“Gagan's technical expertise and leadership are truly exceptional. His deep knowledge of AI systems and product engineering, combined with his ability to manage complex projects, makes him a standout CTO.”
“I'm deeply impressed with Gagan and his team's expertise and dedication they showed working on our project. With their help and support, I could finally see my vision turn into reality.”
Who builds it.
I spent fourteen years scaling engineering teams before I noticed most of the interesting engineering was happening just outside the product — in the seams between tools, inboxes, and spreadsheets where work actually got done.
Buildwithgagan is a deliberately small studio. One person, end-to-end, every engagement. No account managers, no junior handoffs, no "creative director" layer between you and the work. The person on the first call is the person writing the code and the person answering Slack at 11pm when something needs a patch.
I take on two to four projects a quarter. I turn down more than I take — usually because the problem is better solved by a process change than by software, or because a bigger shop is genuinely a better fit. I'll tell you which one you are on the first call.
Before AI swallowed my calendar, I spent a decade leading mobile engineering — shipping native iOS apps in Swift and Objective-C across hundreds of products. That work didn't go away; it's now the second offering here for teams that want the same depth on the App Store side.
BASED
Toronto, working globally
STACK
TypeScript · Python · Swift · Postgres · the usual AI suspects
ALSO SHIP
Native iOS — Swift, SwiftUI, Objective-C
BEFORE THIS
14 years leading engineering at Code Brew Labs
FAVOURITE TOOL
A well-named webhook
WILL NOT BUILD
Cold-outbound spam, deepfake anything
Common questions.
Fixed-scope sprints (4–12 weeks) or fractional CTO retainer (hours/week). Every engagement starts with a 1-week Discovery phase so we understand your problem before writing code.
MVP agent or workflow automation: 4–6 weeks. Production-grade multi-agent system: 8–12 weeks. Fractional CTO: ongoing. You see working software every Friday.
Gagan leads every engagement directly — architecture, code review, and client communication. A trusted bench of senior AI engineers joins for build-heavy phases. No offshored junior teams.
You do. All code, prompts, evals, and infrastructure become yours on delivery. No vendor lock-in, no proprietary wrappers, no license fees.
Scope drives price. Every engagement is shaped to outcomes, not hours. After a 30-minute intro call, you get a fixed proposal with clear deliverables, timeline, and cost. No surprises.
Claude, GPT, and open-source models — we pick for the problem, not the brand. LangChain/LangGraph/CrewAI for orchestration. Python + TypeScript. Infra on AWS/GCP with Kubernetes when scale demands it.
Yes — audit and rescue engagements are roughly 30% of our work. Bring your codebase. We audit, stabilize, and ship what your previous team couldn't.
Recent writing.
How to Build AI Agents Using LangChain
AI Agents are no longer science fiction — they're writing code, querying databases, and making decisions in production right now.
Read essayIndia AI Impact Summit 2026 Coverage
India just hosted the biggest AI summit the world has ever seen. Feb 16–21 | Bharat Mandapam, New Delhi. 100+ countries | 20+ heads of state.
Read essayLet's build something.
Send a short note about what you're trying to automate, and I'll reply within a day with a 30-minute call or a fixed-price audit proposal — whichever fits.
Free 30-minute audit for qualified companies
Fixed-price quote after the first call — no estimate drift
NDAs, DPAs, and VPC-only builds available on request
Currently taking 2 new projects this quarter