Our Bansar Technologies profile turned up a puzzle: $141 million and 534 Texas state contracts, but almost no visible H-1B or green-card sponsorship under Bansar's own name. We went looking for where the workers actually come from — not by guessing, but by mining the filenames Texas agencies themselves attach to contract paperwork, then cross-checking those names against the state's own contract subject lines. Three confirmed cases later, the answer isn't "the employees don't exist." It's that the same named individual keeps doing the same job at the same agency while the company billing the state for their time changes underneath them.
Every Texas state contract in the LBB Contract Management System (cms.lbb.texas.gov) can carry file attachments — purchase orders, change notices, and the like. Agencies aren't consistent about how they name these files, but several agencies' naming conventions accidentally embed the placed worker's name directly in the filename: a 2026 Railroad Commission/Bansar purchase order was literally named 26-1346 -Bansar - ITSAC - SGOGINENI.pdf. We built a scraper that walks every contract page for a given vendor, reads whatever attachment filenames are listed (no PDFs opened, no OCR — just the filename text Texas's own system already displays publicly), and mines each one for a name-shaped candidate.
Run across 14 vendors and 984 contracts, that surfaced roughly 18 plausible individual names once administrative-label noise ("Staff Aug," "POCN #4," "Solicitation") was filtered out. We then searched those exact surnames against the full, unfiltered Subject field of every Texas state contract on file — not just the 14 vendors we'd scraped — which is how the three swaps below turned up. Several apparent hits were excluded as false positives: "Neerukonda," "Manukonda," "Kammari" (as a stand-alone token), and "Sankanna"/"Kannam" all share a substring with a name on our list but belong to different people with genuinely different surnames.
In each case below, the state awarded a brand-new, separate contract to a different company — not an amendment, not a subcontract, a new contract number and a new vendor — naming the identical worker for the identical role at the identical agency.
| Vendor | Contract # | Subject (as recorded by the state) | Award Date | Value |
|---|---|---|---|---|
| Loblolly Consulting LLC | 22-00023 | Information Technology Staff (ITSAC)_R. Konda | 2021-09-01 | $210,000 |
| Bansar Technologies Inc | 24-40268 | ITSAC Bansar Contractor_ R. Konda | 2024-01-02 | $154,550 |
| Bansar Technologies Inc | 25-00026 | ITSAC Bansar Contractor_ R. Konda | 2024-09-01 | $224,000 |
| Bansar Technologies Inc | C2600032 | FY26 ITSAC ITD .BANSAR TECH contractor R. Konda | 2025-09-01 | $224,000 |
Matches our TABC investigation's independently-sourced finding (signed purchase-order review) that the same individual, "Raja Konda," billed as QA Manager across five fiscal years, first through Loblolly Consulting then through Bansar Technologies.
| Vendor | Contract # | Subject (as recorded by the state) | Award Date | Value |
|---|---|---|---|---|
| Allied Consultants | 20-1014 | IT Contract: Jag Nanugonda | 2020-03-11 | $109,450 |
| Allied Consultants | 21-0142 | ITSAC: Jag Nanugonda | 2020-09-01 | $109,450 |
| Bansar Technologies Inc | 22-0109 | ITSAC Contract: Jag Nanugonda | 2021-09-01 | $193,764 |
Newly identified by this report. Bansar's own attachment filenames at Railroad Commission separately named three more individuals (Narasimha Tirumala, Praveen Vemanedi, Ramarao Kavuru) with a recurring "VW" suffix whose meaning we have not resolved — possibly initials, possibly an unrelated internal code.
| Vendor | Contract # | Subject (as recorded by the state) | Award Date | Value |
|---|---|---|---|---|
| 22nd Century Technologies Inc | CA-0001994 | IT-L2 SoftwareDev-RajKammari | 2021-09-27 | $725,120 |
| Nipun System Inc | CA-0005930 | BTR Stfwr Dev2 RajKammari | 2023-12-08 | $662,640 |
Newly identified by this report, and notable because Bansar isn't involved at all — this is the same mechanism operating independently between two other companies, which argues the pattern is structural to how Texas staffs these contracts rather than specific to any one vendor. One caveat: $662-725K is well above the per-worker rate implied by the other two cases, so this contract may cover more than a single individual's billing — Kammari may be the named lead on a larger staffing line item, not its sole worker.
Subcontracting would mean one company holds the state contract while quietly paying a second company to actually deliver the work — the subcontractor's name might never appear in public procurement records at all. That is not what the data shows here. In every case above, the second company holds its own separate, independently-awarded contract with the state. Nothing links the two contracts on paper except the worker's name recurring in the subject line. This is vendor succession: as one company's contract term ends, a competing company wins a new one for the same position, and the incumbent worker moves with it.
That distinction is what actually answers the "where are the employees" question. LCA and PERM disclosure data only capture new sponsorship activity — a worker who already has status (a green card, a pending petition transferred rather than newly filed, or citizenship) generates no new federal filing just because the staffing company invoicing for their time changes. A company can run $141 million in state contracts on a comparatively small, continuously-recycled worker pool and still show almost no sponsorship footprint under its own name, simply because most of that pool was never newly sponsored by that company to begin with.
What remains genuinely open is whether these "competing" vendors are independently owned. Benign labor-market churn — an incumbent contractor naturally following whichever firm wins the next bid cycle — and coordinated bid rotation among commonly-owned or affiliated shell entities would look identical in this data. Resolving that requires Texas Secretary of State registered-agent and officer records, which are not bulk-available (SOSDirect charges per document) and which we have not yet pulled for any of the five vendors named in this report.
This investigation is produced independently — one person, 23 years of data work, living on $1,419/month Social Security. If this work matters to you, a small donation makes a real difference.
♥ Donate via Square More Reports